UK Vintage Radio Repair and Restoration Powered By Google Custom Search Vintage Radio and TV Service Data

Go Back   UK Vintage Radio Repair and Restoration Discussion Forum > General Vintage Technology > Components and Circuits

Notices

Components and Circuits For discussions about component types, alternatives and availability, circuit configurations and modifications etc. Discussions here should be of a general nature and not about specific sets.

Closed Thread
 
Thread Tools
Old 12th Jun 2011, 10:55 pm   #21
neon indicator
Retired Dormant Member
 
Join Date: Jun 2010
Location: Co. Limerick, Ireland.
Posts: 1,183
Default Re: Russian Military Subminature "pencil" valves

I've ordered the following
1j24b = Very low heater power pentode heater 12mA/15mW (Anode 0.8mA 600uS)
1j37b = dual g1 pentode! Mixer heater 60mA/70mW (Anode 2mA)
1p24b = 4W pentode RF or AF power. I think

The 1j18b I bought already is 24mA 1.2V (NOT 1.5V) heater and works usefully down to 15V (so 2x PP3 possible) or up to 100MHz oscillator (42v HT).

I have some cheap LT1073 now and MUR120 fast diodes so can make battery powered 15V to 90V PSU from 1.5V or 6V battery in. Also 1.2V from 6V in. The LT1073 is only a bipolar switch PSU at 19kHz. Modern chips are typically FET and 1MHz to 3MHz switching, but the LT1073 has these advantages[list][*]Cheap on eBay[*]DIL, not SM[*]up to 45V output directly (90V by adding "doubler diode" to basic "boost" design)[*]Buck, Boost, Buck-boost, Ćuk or SEPIC converter according to wiring

http://en.wikipedia.org/wiki/Buck_converter
http://en.wikipedia.org/wiki/Boost_converter
http://en.wikipedia.org/wiki/Buck-boost_converter
http://en.wikipedia.org/wiki/%C4%86uk_converter
http://en.wikipedia.org/wiki/SEPIC_converter

The PSU(s) need to be in a metal box with L C filter on input and outputs. Veroboard/stripboard is fine. I make metal boxes out of the round Lidl Expresso Coffee tins, the black "paint" is solderable, by cutting them up with scissors!

Interestingly if you use the doubler Boost circuit for 90V in Linear Technology data sheet, but for 45V, then the intermediate is 22V, just right for the VFD seven segment panels Anode/Grid drive. I'm experimenting with connecting the VFD filament ends to +5V supply and driving it in bridge with a TDA1308 stereo headphone amp via a pair of capacitors for the 4.5V filament AC power.

See again http://www.radiomuseum.org/forum/rus...ure_tubes.html
neon indicator is offline  
Old 29th Jun 2011, 9:11 am   #22
neon indicator
Retired Dormant Member
 
Join Date: Jun 2010
Location: Co. Limerick, Ireland.
Posts: 1,183
Default Re: Russian Military Subminature "pencil" valves

I now have Russian data sheets for ALL the common valves and some English Translations.
(links to datasheets http://www.techtir.ie/blogs/watty/miniature_valves_2)

I'm still not sure of the value of the 1j17b (60mA filament) or the 1j29b (centre tap 2.4V @ 30mA, 1.2V @ 60mA) Pentodes. I'm looking for someone that does Russian!

However the following people will find generally helpful read any scanned datasheet in Russian:
сетка = grid
анод = Anode
экран = Screen or Shield
катод = Cathode
накала = filament
минус = Minus / Negative
плюс = Plus / Positive
ток = Current
напряжение = Voltage
первый = first
второй = second
треть = third

i.e. Voltage G2: напряжение сетка второй

Data sheets on some parts give 50, (!), 500 and 1000 Hr rating. I'm presuming like light bulb filaments the life time is tiny when over-run by 10% and absolutely huge under run by 20%

Since ultimately the Anode current must be all coming from the Cathode (in this case the filament), then the importance of de-rating to get 10,000 to 20,000 life is not just filament voltage but Anode / Operational point too. Most seem to still give useful operation as low as 0.6V filament or 1.1V filament and 18V HT.

Looking at curves most them appear to still be very good at 48V HT and 1.0V to 1.1V filament, so this is what I shall test at. On hybrid designs I'm thinking a transistor regulator for filaments and on "pure" valve designs the old 2V valve scheme of series resistor. I think 1 or 2 resistors and direct to cover the main parts of NiMH discharge curve (1.25V typical, 1.35V very fresh, 1V near end) . It means the filaments should never be connected while battery charging (Over 1.48V).

As far as I can figure the common ones are:
  • General purpose Pentode, 1Ж18Б = 1j18b rated to 60MHz, but may be usable over 100MHz. G3 internally connected to Cathode/Filament, limits RF applications. These are cheapest and most available. Medium filament at 24mA. Will work as a triode using G2 as the Anode.
  • More general purpose Pentode?, 1Ж17Б = 1j17b Suitable for more RF circuits as G3 is and shield/screen are on separate wires, but high current 54mA filament. Slightly higher power and emission at low due to higher current Filament.
  • Driver and low power output Pentode. 1Ж29Б = 1j29b. Centre tap filament to allow 30mA @ 2.4V or 60mA @ 1.2V.
  • Ultra low power RF / AF Pentode 1Ж24Б = 1j24b. Filament only 17mA @1.2V. G1, G2, G3 all on separate wires. Not as cheap.
  • Amazing 2W Pa power RF / AF Pentode 1П24Б = 1p24b.Centre tap filament to allow 30mA @ 2.4V or 60mA @ 1.2V. G1, G2, G3 all on separate wires. Usually a bit more expensive, not so many sources. In Class C or push pull Class B, up to 4W at 30MHz or 4W Audio is possible. Life may only be 500 Hrs at full power at 1.3V filament. Suggest 1.2V max filament and 1W Pa max. Up to 150V HT. Can be used 12V HT at lower power.
  • Mixer, frequency doubler, RF/AF full wave rectifier, 1Ж37Б = 1j37b. This unique pentode is not a hexode or a bi-grid valve. It's a pentode with two G1 (control grids). The unique multi-rod construction of the anode and grids instead of perforated foil or helical fine wire means the G1 rods can be in two identical sets. It's analogous to a dual gate depletion FET. One g1 can be g and the other "plate" for "gammatron" operation. Earth the other electrodes.

There are two pin out versions of the 1j29b, older has g1 on middle pin. The 1Ж29Бp (1j29b-v? ) has the 7 base wires arranged like B7G, no middle wire.

The base wiring diagrams are confusing till you realise they count cut off wires used for internal support/connection. e.g. 1j18b diagram shows 6 wires, but 3 and 4 don't exist, or are very short cut off stubs. So don't assume there are "snapped off" wires

Today I will see if 1j37b is simple to use not just as RF/AF full wave rectifier, frequency doubler and mixer, but as tunable and crystal Oscillator / Mixer. The data sheets label control grids as g1' and g1'', I will use g1a = pin/wire 2 and g1b as pin/wire 5

Some models have an additional Shield (screen in Russian) on a separate wire as well as G1, G2 and G3 on separate pins. This is shown diagrammatically as near the filament and is earthed in schematics. Any thoughts on it?

Conclusions so far:
  • These are very readily available and can be from 50p to £1 each including postage.
  • Lifetime may be as good as any valve if run closer to minimum spec (which is 0.9V filament) and care is taken that Anode current is not too high on filament, which will be helped if both ends of filament are RF/AF ground on higher power (at low power pre-amp for RF etc, RF earthing grid and RF chokes or transformer on both filaments is fine). At the 1.4V filament and max rated Anode current/volts the life is 95% live to 500 hrs
  • Filament should be below 1.2V, these are not like traditional D series 1.5V valves.
  • HT on all except 1p24b as RF transmitter should be 48V nominal max, not more than 56V (fresh battery). 18V operation is feasible.
  • There are variations of pin-out and the black spot for pin1 isn't always correct or even existing, but examination and final check with meter for filament resistance will verify.
  • Should be possible to replicate old circuits back to 1920s but with lower voltage, or 1950s battery portables or "more modern" designs.
  • A modest headphone amp is possible or a more ambitious 1W (2 x 1p24b) to 2W (4x1p24b) audio amp using small mains transformer with two 110V windings in push pull.
  • Possibly using 4 x 1p24b in 2x push pull 8W or 70MHz transmitter might be possible. About 140V HT. At that power lifetime might not be great! (CW, SSB 8W, AM 2W carrier no modulation. FM could be 8W but might need de-rated for life extension! PSK31 and SSTV should likely be de-rated (linearity on PSK31, life on SSTV)
  • lots of existing schematics for these and lots of room for experiment.
  • On all except 1p24b the grid /filament- can be zero or even +0.5 with a 1.1V filament! Especially at lower HT volts.

Thoughts?

Last edited by neon indicator; 29th Jun 2011 at 9:16 am.
neon indicator is offline  
Old 29th Jun 2011, 5:03 pm   #23
julie_m
Dekatron
 
Join Date: May 2008
Location: Derby, UK.
Posts: 7,735
Default Re: Russian Military Subminature "pencil" valves

Quote:
Originally Posted by neon indicator View Post
Mixer, frequency doubler, RF/AF full wave rectifier, 1Ж37Б = 1j37b. This unique pentode is not a hexode or a bi-grid valve. It's a pentode with two G1 (control grids). The unique multi-rod construction of the anode and grids instead of perforated foil or helical fine wire means the G1 rods can be in two identical sets.
This sounds like an OR gate: both G1a and G1b must be negative to cut off the anode current. As opposed to a hexode / heptode, which works like an AND gate: either G1 or G3 going negative will cut off the anode current.

Could you still use this with a separate oscillator, to build the usual type of frequency changer? Would there be a problem with inter-G1 capacitance causing the oscillator to be pulled off-frequency?
__________________
If I have seen further than others, it is because I was standing on a pile of failed experiments.
julie_m is offline  
Old 29th Jun 2011, 5:49 pm   #24
neon indicator
Retired Dormant Member
 
Join Date: Jun 2010
Location: Co. Limerick, Ireland.
Posts: 1,183
Default Re: Russian Military Subminature "pencil" valves

It's apparently designed to be a mixer OR a mixer osc.
I've found an Engineering Russian Native in Ireland (Mech rather than Electronic)

I made mistake reading 1p24b data, it's 80mA per filament @ 1.16V, so 160mA total, actual tests. Datasheet is actually 230mA.

The 1j37b sheet seems to be 2.25pF g1' or g1'' (g1a and g1b) with screen can on glass

Actual measure (with my PIC based meter and valve cold)
g1a 2.6pF (all other wires tied)
g1b 2.6pF (all other wires tied)

g1a to g1b and all other wires "floating" = 1.5pf approx
The meter is accurate to +/- 0.1pf
neon indicator is offline  
Old 29th Jun 2011, 6:07 pm   #25
neon indicator
Retired Dormant Member
 
Join Date: Jun 2010
Location: Co. Limerick, Ireland.
Posts: 1,183
Default Re: Russian Military Subminature "pencil" valves

You don't often get this in a data sheet

1p24b 10mm x 40mm "fluted" glass tube
Allegedly able for 1.5W anode dissipation and 4W per valve Class AB using 180V HT
(2W seems more reasonable O/P power per valve in Class AB for any sort of life!)

Maximum single impact acceleration (up to 10) = 500g
Maximum repetitive impact acceleration (up to 4000) = 150g
(g I presume is 9.8m/s/s)

Also a table of peak ratings for 50 hour life!
400V Anode and 800mA current, 3uS pulse
(miniature MIG HF Radar or what?)
neon indicator is offline  
Old 29th Jun 2011, 7:29 pm   #26
murphyv310
Dekatron
 
murphyv310's Avatar
 
Join Date: Sep 2006
Location: Kilmarnock, Ayrshire, UK.
Posts: 5,420
Default Re: Russian Military Subminature "pencil" valves

As yet I haven't started to do anything with mine as I have a fair bit of other things to do but keep a look out.
__________________
Cheers,
Trevor.
MM0KJJ. RSGB, GQRP, WACRAL, K&LARC. Member
murphyv310 is offline  
Old 29th Jun 2011, 11:13 pm   #27
julie_m
Dekatron
 
Join Date: May 2008
Location: Derby, UK.
Posts: 7,735
Default Re: Russian Military Subminature "pencil" valves

Quote:
Originally Posted by neon indicator View Post
It's apparently designed to be a mixer OR a mixer osc.
Of course! I think I see how it works now You use one of the two G1s as the oscillator grid, have the IF transformer in the anode as per usual, and you feed the antenna signal into the other G1. As long as just enough of the oscillator frequency (which = station + IF) can break through the anode load, the oscillator will keep going at its own frequency determined by one gang of the tuner. Only, the antenna signal is also gating the oscillator (the electron stream is repelled when both inputs are negative-going but allowed through any time the antenna grid is positive-going), so you get sum and difference frequencies. And the difference, which of course is the IF, gets coupled through the transformer into the secondary and on to the next stage.

By genius! It only works with that exact geometry, as well. No way you could do it with Mullard-type spiral grids .....

Only thing now is, how do you do the IF stage? I don't suppose they make any variable-µ valves with the grid rods in a sort of V-shape, do they?
__________________
If I have seen further than others, it is because I was standing on a pile of failed experiments.
julie_m is offline  
Old 30th Jun 2011, 8:30 am   #28
neon indicator
Retired Dormant Member
 
Join Date: Jun 2010
Location: Co. Limerick, Ireland.
Posts: 1,183
Default Re: Russian Military Subminature "pencil" valves

The 1j18b, 1j29b and 1j17b are all in different Russian Military sets as IF amps.

See http://www.radiomuseum.org/m/military_su_en_1.html
I have no account there, so can't download any of the Russian Military Radio schematics

From what circuits I have seen and testing with the 1j18b I think you can do AGC by voltage on g2?

This is the best set of articles I have found http://www.radiomuseum.org/forum/rus...ure_tubes.html

There are loads of German sites with schematics and complete design/discussion of 1920s and "modern" Radio (one with VHF FM and stereo decoder!) as well as I few Russian Almost nothing in English.
I will add any more links I find to the list here http://www.techtir.ie/blogs/watty/miniature_valves_2

This 1969 Military vehicle RX/TX has astounding number of these Pencil-valves:
http://www.radiomuseum.org/r/unknown_r_130p_13.html
Power type and voltage Storage Battery for all / 26 Volt
Loudspeaker - For headphones or amp.
Model: R-130 {Р-130}
Material Metal case
Shape Miscellaneous shapes - described under notes.
Dimensions (WHD) 17 x 10 x 12 inch / 432 x 254 x 305 mm
Notes AM/CW transceiver, range 1.5 MHz to 10,99 MHz with 10 preset channels. Mainly used in soviet and her allies armed forces in car and combat vehicles.
Tubes: 4× 1p24b, 32× 1j29b, 25× 1j24b, 2× gu-50. R-130M is never version.

4 x 1p24b is likely RF driver, , or maybe 1 pair is headphone drive. 200mA heater each
2 x gu-50, like LS50, a RF power pentode. Pushpull RF PA driven by pair of 1p24b? Not a pencil valve.
25 x 1j24b likely RF amps, oscillators, AF preamps. Only 17mA heater each
32× 1j29b likely IF amps, AF amps, drivers. 60mA heaters
neon indicator is offline  
Old 2nd Jul 2011, 10:21 pm   #29
neon indicator
Retired Dormant Member
 
Join Date: Jun 2010
Location: Co. Limerick, Ireland.
Posts: 1,183
Default Re: Russian Military Subminature "pencil" valves

Some tests on the 1p24b RF PA.

2.2 Ohms x 2, one in each f- leg gives just over 1.1V from approx 1.3V to ground. f+ to NiMH +. About 110mA + 110mA max, typical 100mA, so 200mA to 220mA from battery.

lower battery then add 2 more 2.2 to give 1.1 Ohms
and then as battery lower, (0.95V to 1.1v) connect direct.

with G1 = 0V (actually f+ less 1.3V and f- less 0.2V, so sort of -0.65 average)
Code:
  5   9   10  12 15 17.5 18 20 25 27.5  30  32 35  40    45
0.35 1.1 1.4 2  3   4    4.4 5  7.4 8.6  9.9 11 12.5 15.5 18
with G1 = -5V (actually f+ less 1.3V and f- less 0.2V, so sort of -5.65 average)
Code:
  5   9   10  12 15 17.5 18 20   25  27.5  30  32 35   40    45
  0   0   0    0   0   0    0  0.02 0.2 0.33 0.8  1  1.44 2.66  4.14
with G1 = f+
Code:
  5    10  
 11    35
With g2 = anode, g3 = ground

A serious Audio PA with 2 x push pull and +44V HT
scary RF power for such a small thing at recommended 135V!

Very likely useful with single PP3 supply and bias between f+ and f- volts. I'll see how well it works 9V, 12V and 18V for Audio, RF amp and Osc sometime after building the +44V HT push pull audio amp.

I think for class AB RF PA at +135V HT a -9V or more grid bias may be needed. Should be about 1uA to 5uA? I think I want between 3mA and 7mA anode + g2 bias current at +135V HT for RF Class AB, since at 45V HT, 0 to -5 is 18mA down to 4mA approx.

Last edited by neon indicator; 2nd Jul 2011 at 10:30 pm.
neon indicator is offline  
Old 3rd Jul 2011, 8:42 am   #30
neon indicator
Retired Dormant Member
 
Join Date: Jun 2010
Location: Co. Limerick, Ireland.
Posts: 1,183
Default Re: Russian Military Subminature "pencil" valves

Obviously the above will be a triode like curves. The Datasheets have the Pentode curves, which require g2 to be at a fixed voltage. G2 current will be "high" and then drop as anode voltage increases.

Since you can't easily do negative feed back via cathode resistor (direct filament) an ultra-linear type PP transformer for a pair of 1p24b and aux windings to screen would be interesting.
neon indicator is offline  
Old 3rd Jul 2011, 11:05 am   #31
neon indicator
Retired Dormant Member
 
Join Date: Jun 2010
Location: Co. Limerick, Ireland.
Posts: 1,183
Default Re: Russian Military Subminature "pencil" valves

I added graph of the above readings here http://www.techtir.ie/blogs/watty/miniature_valves_2
(the forum here doesn't seem to allow inline images. Or else I'm doing it wrong)

A search with Google or ebay finds PLENTY of these type valves and as low as 50p each in quantities of 100! About €1 to €1.50 each per 10 or 12 lot inc postage.
neon indicator is offline  
Old 4th Jul 2011, 2:58 pm   #32
neon indicator
Retired Dormant Member
 
Join Date: Jun 2010
Location: Co. Limerick, Ireland.
Posts: 1,183
Default Re: Russian Military Subminature "pencil" valves

More study of Russian Data sheets and my own measurements:

1j24b is indeed only 11.6mA @ 1.1V filament. (recommended current is 11 to 13). So putting a 22 Ohms in series with NiMH battery is advised. An Alkaline would need maybe 47 Ohms when fresh!

With an HT of 44V, G2= HT, Anode Load = 56K, G3 = 0V (earthy end of the 22 Ohms on f- is LT and HT common, LT+ direct to f+) and G1 = 0V via 100K Ohms.

HT+ current less than 0.4mA! (g2 + anode)
Voltage Gain a bit more than 17x
2V peak to peak and 0.2V peak to peak square waves at 1kHz are same gain and look good on output.

Putting 220k in series with G2 (instead of direct to HT+) causes slight negative overshoot and reduces gain to 10. (no decoupling). No doubt is more linear?

Connecting G2 to Anode (Triode Mode) reduces gain to about x5, but loading may be wrong (likely 56K is too high)

With greater understanding of the Russian Data sheet curves (and confirmation of testing) I suggest:
Approx Anode loads min Ohms @ approx 45V HT Class A (filament)
1j17b 33k (54mA) Suitable very high RF and grounded grid. Separate Shield and g3
1j18b 47k (24mA) Simple connections. No external G3 connections.
1j24b 56k (12mA) Suitable very high RF and grounded grid. Separate Shield and g3
1j29b 12k (60mA) RF driver 1j29br (1Ж29Бр) version recommended
1j37b 39k (54mA) Unusual Dual G1, mixer/osc/detector
1p24b 2.7k (200mA) 2W+ PA to maybe 150MHz

I'm no valve expert, but those are reasonable starting points with 40V to 50V HT.

I suggest the 1j17b is not worth purchasing as it's replaceable by 1j24b (lower power 11mA heater) or higher spec 1j29b (60mA filament). Note there are two versions of the 1j29b. The earlier one has a 7th middle wire, the later ones have all wires on edge like a B7G (except smaller!). The later one is marked 1Ж29Бр (In English maybe 1j29Br or 1shBr. (There is a -v version listed http://www.radiomuseum.org/tubes/tube_1j29b-v.html but I not yet seen one). The 1j29b is nearly x5 the filament power (60mA vs 11.3mA) of 1j24b but also is x4 max anode current, 5.5mA vs 1.35mA, so if you need the extra Anode power (55mW static class A vs 15mW class A) the 1j29b not much less efficient. The 1j29b will run off a 45V or 90V HT and 1j24b suited to a 45V HT

Most of the valves appear to be for 45V HT as G2 max is 60V and recommended is 45V (A military pack would maybe be 60V peak, 45 nominal to 39V near flat)
The exceptions are the 1j29b which can be 90V and the 1p24b which can be 140V (to 500V pulse operation).

The 1j24b works with only 4.5uA grid current with G1 about +0.65 above f+. Since f+ is at 1.35V and f- at 0.25. This gives Anode current of 0.93mA @ 32V HT.
With G1 at 0v the Anode current is 0.6mA
Datasheet Max recommended Ia is 1.5mA (possible HT +45V and higher at G1 = 0V)

At +45V HT and class A, all can run with G1 = 0V, f+ = 1.35V, f- = 0.25V, except the 1p24b.

For "push pull" class AB, you need -5V from somewhere for the 1p24b with HT=45, and something between -9V and -12V (not tested yet) for +135V HT

The minimum HT to do anything is +15V to +18V (more realistic is +18V), except the 1j29b and 1p24b.

The 1j29b will operate at +9V. You may need g1 slightly positive, but still less than f+ O/P could be 30mW using class AB push pull, or about 10mW single ended.

The 1p24b will operate with "TTL logic levels" off a 5V supply. with G1 at 0V there is no current. At g1 = f+ there is 12mA! At +9V the 1p24b can do about 35mW SE and a pair in Class AB maybe over 100mW. At 44V HT maybe 800mW to 1W and at 140V HT perhaps over 8W at over 100MHz in push pull.

Summary (brackets is filament @ about 1.1V) for 45V nominal HT, minimum class A loads:
Ignore 1j17b
1j18b 47k (24mA) Simple connections. No external G3 connections. No external connection to "Shield" Cheap.
1j24b 56k (12mA) Suitable very high RF and grounded grid. Separate Shield and g3
1j29b 12k (60mA) RF driver
1j37b 39k (54mA) Unusual Dual G1, mixer/osc/detector
1p24b 2.7k (200mA) 2W+ PA to maybe 150MHz. No -ve bias needed for class A at 45V. -5V for class AB @ 45V.

All loads assume class A and
f+ = +1.3V approx
f- = + 0.2V approx (series resistor to 0V)
LT- and HT- are common 0V rail
g1 biased by 100k Ohms to 0V (higher may be fine), about -0.2V compared to f-
You have f- and f+ correct polarity!
g2 to +HT 45V
g3 (and Screen/Shield if available) to 0V (except 1j18b)

The 1p24b seems to be heavier glass, ribbed maybe Quartz for power handling?
neon indicator is offline  
Old 4th Jul 2011, 6:25 pm   #33
neon indicator
Retired Dormant Member
 
Join Date: Jun 2010
Location: Co. Limerick, Ireland.
Posts: 1,183
Default Re: Russian Military Subminature "pencil" valves

Ok made mistake on my scope settings on the 1j24b circuit.
f+ = +1.3V approx
f- = + 0.2V approx (22 Ohms series resistor to 0V)
LT- and HT- are common 0V rail

g3 and shield to 0V
g1 to 0V via 100K

input and output decoupling capacitors 220nF

Ia = 270uA (0.27mA)
HT = 44V
R anode load = 68K (DC anode volts = 24V)

With Screen (g2) to HT gain is 30. Slight rounding of square wave
With series resistor G2 to HT of 100 K Ohm, gain is 25 and square wave looks good. As you increase R g2-HT the gain drops but you get overshoot 220K is gain 20, 560K is 50% overshoot.

So just a bit less than x2 Anode load seems to be limit.
neon indicator is offline  
Old 4th Jul 2011, 10:24 pm   #34
neon indicator
Retired Dormant Member
 
Join Date: Jun 2010
Location: Co. Limerick, Ireland.
Posts: 1,183
Default Re: Russian Military Subminature "pencil" valves

Basic Tests now done on 1j29b "medium" power RF/IF/AF Pentode.

It needs 2 x 10 Ohm resistors on the pair of f- connections to LT- / 0V and connect common f+ to NiMH battery LT+. An alkaline cell may need a larger pair of resistors. An additional pair of 10 Ohm and then short could be added as the battery gets low.

Additional Triode Mode graph http://www.techtir.ie/blogs/watty/miniature_valves_2 added, the Datasheet's Pentode curves are adequate.

The 1j29b is 54mA to 60mA filament (best keep to 54mA in class A) and likely works to over 150MHz. (to be tested). It can be used RF/IF/AF and may be needed as driver for a pair of 1p24b in ClassAB giving nearly 8W.

Further experimenting to see how well it works as headphone or speaker amp on a 1 valve radio in class A on 9V, 18V or 27V and Class AB Push pull.
neon indicator is offline  
Old 5th Jul 2011, 6:29 pm   #35
neon indicator
Retired Dormant Member
 
Join Date: Jun 2010
Location: Co. Limerick, Ireland.
Posts: 1,183
Default Re: Russian Military Subminature "pencil" valves

Anyone good at Russian?

http://www.radiostation.ru/home/sterzhni/01.html
http://www.radiostation.ru/home/sterzhni/02.html
http://www.radiostation.ru/home/sterzhni/03.html
http://www.radiostation.ru/home/sterzhni/04.html
http://www.radiostation.ru/home/sterzhni/05.html
http://www.radiostation.ru/home/sterzhni/06.html
http://www.radiostation.ru/home/sterzhni/07.html
http://www.radiostation.ru/home/sterzhni/08.html
http://www.radiostation.ru/home/sterzhni/09.html
http://www.radiostation.ru/home/sterzhni/10.html
http://www.radiostation.ru/home/sterzhni/11.html

Scans, so Google Translate doesn't work!

I can only get a voltage gain of 7 from the 1j29b, it seems to like 3k9 Anode load and 4k7 g2 load with 45V HT
neon indicator is offline  
Old 5th Jul 2011, 8:23 pm   #36
neon indicator
Retired Dormant Member
 
Join Date: Jun 2010
Location: Co. Limerick, Ireland.
Posts: 1,183
Default Re: Russian Military Subminature "pencil" valves

The "basic" 1j29b has 7th wire (g1) in the middle. The 1j29b-r (marked 1Ж29Бр ) is reputedly rated 5,000hrs and is similar gain (voltage gain 8 in Pentode with screen feedback), but a wire layout like B7G, no middle wire. Completely different pinout.

Using 5k6 Ohm Anode load the Va is 21.5V for 44.6VHT. The G2 volts is 42, with a 6k8 Ohm load.

Using +1.34V NiMH (measured on load) (on wire 6) and 10 Ohms on each f- (wires 2 & 5) the filaments are 25.9 and 26.1 (calculated from volts on the 1% 10 Ohm resistors) total current 52mA.

The 6.8 K on g2 rather than direction connection adds negative feedback and makes gain more immune from supply changes and gives good square wave response. Tested 200mV and 2V peak to peak drive. Very little change between 30V and 45V HT
neon indicator is offline  
Old 5th Jul 2011, 9:27 pm   #37
neon indicator
Retired Dormant Member
 
Join Date: Jun 2010
Location: Co. Limerick, Ireland.
Posts: 1,183
Default Compare 1j29b with DL96 (or DL91).

Looking at conventional Battery valves the DL96 seems very broadly similar (50mA @ 1.4V filament and about 200mW max output, 90V or 67V HT).

Has anyone tried DL96 or DL91 at 45V HT?

Anyone know what the voltage gain at 45V HT for DL91 or DL96 might be with 4k7 Ohm load?

I think I should be running the 1j29b and 1j29b-r at quiescent of about 2.5mA to 2.75mA which would be closer to a 10K Ohm load. But I'd need some negative bias for that. With G1 = 0V, the 22V Anode with 44V HT is about 5.6K, about 3.9mA. Max Ia is 5.5mA. I think I need about -1 to -1.5V looking at Datasheet.

Last edited by Brian R Pateman; 5th Jul 2011 at 9:54 pm. Reason: Shouty format removed.
neon indicator is offline  
Old 5th Jul 2011, 11:36 pm   #38
neon indicator
Retired Dormant Member
 
Join Date: Jun 2010
Location: Co. Limerick, Ireland.
Posts: 1,183
Default Re: Russian Military Subminature "pencil" valves

Well, the 3.9mA is obviously too high and 5.6K too low. Original look at load line I did on the Pentode curve told me 10K and 2.5mA

I used -5V and 390K to the g1, and left existing 100k to 0V, this gives -0.95 on g1.

Anode load to 44.5V HT is now 10K for 22V anode volts (no signal). Screen grid G2 at 15K Ohms and 41V (About 0.23mA or 230uA) The "rule of thumb" for 1j24b and 1j29b anyhow seems to be screen grid "load" = 1.5x the Anode load for good transient response and some negative feedback.

Vout/Vin gain is now 10. If g2 connected to HT (44.5) gain is 12. I think some local negative feedback mitigates battery discharge. With 15k on G2 there is little change in gain from 34V to 44V, with G2 = HT the gain changes from about 10.5 to 12 from 34V to 44V

10K and 22V Va with 44.5V HT is about 2.25mA (G1 = -0.9V via 390K & 100K divider on -5V).

If using this to drive a 1p24B or pair of them, you'll need a grid bias supply anyway. If using it with a pair of PP3s (16V to 18V HT) then the g1 can be at 0V.
neon indicator is offline  
Old 6th Jul 2011, 8:10 pm   #39
neon indicator
Retired Dormant Member
 
Join Date: Jun 2010
Location: Co. Limerick, Ireland.
Posts: 1,183
Default Re: Russian Military Subminature "pencil" valves

Realisation today that the 1j29b can be biased (-1.1V) by running its filament in series with a pair of paralleled 1j18b off 2 x NiMH. Stick 330 ohms across the 2 x 1j18b which have 22 Ohms in series each at f- to ground.

It was studying this schematic... http://www.hi-ho.ne.jp/ux-45/schmatic/SMT4V1.gif

Also an RF stage with choke in series with f+ & f- and decoupling can be in series with 2nd valve same rating.

Similarly if you ran parallel / series and had 220mA total you could put one 1p24b in series biased to -2.4V as the "top" of the chain off 3 x NiMH

I'm not used to Valves or the direct heaters, The 1j18b, 1j24b definitely need Vg1 = 0
neon indicator is offline  
Old 6th Jul 2011, 8:39 pm   #40
murphyv310
Dekatron
 
murphyv310's Avatar
 
Join Date: Sep 2006
Location: Kilmarnock, Ayrshire, UK.
Posts: 5,420
Default Re: Russian Military Subminature "pencil" valves

My little audio amp is totally conventional, the filaments are connected to a 1.5v DC supply the anode and screen are fed from a 160v supply the screen is fed via a 56k resistor and decoupled by a 0.01uf to common (Filament +) The anode is fed via a 150k from 160v giving 120v on the anode. the grid is returned to the filament - by 4.7megs a 1000hz input is fed to the grid via a 0.001uf cap from a signal generator the anode is output in the conventional way via a 0.01uf cap. 100mv input gives me 8 v p-p but any higher input results in clipping, not symmetrical though which makes me think the bias is out.
__________________
Cheers,
Trevor.
MM0KJJ. RSGB, GQRP, WACRAL, K&LARC. Member
murphyv310 is offline  
Closed Thread

Thread Tools



All times are GMT +1. The time now is 11:07 pm.


All information and advice on this forum is subject to the WARNING AND DISCLAIMER located at https://www.vintage-radio.net/rules.html.
Failure to heed this warning may result in death or serious injury to yourself and/or others.


Powered by vBulletin®
Copyright ©2000 - 2024, vBulletin Solutions, Inc.
Copyright ©2002 - 2023, Paul Stenning.