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 > Specific Vintage Equipment > Vintage Amateur and Military Radio

Notices

Vintage Amateur and Military Radio Amateur/military receivers and transmitters, morse, and any other related vintage comms equipment.

Closed Thread
 
Thread Tools
Old 1st Nov 2019, 4:43 pm   #1
JohanBee
Triode
 
Join Date: Apr 2014
Location: Stockholm, Sweden
Posts: 17
Default 6BE6/EK90 vs 6CS6/EH90 as frequency mixer in Trio 9R59DE ?

Hi, I recently bought a Trio 9R59DE in very good condition as my winter project to play with. I know of the rule ”if it ain’t broke don’t fix it”, but the circuits solutions are rather crude and seems copied from others, or made by not very experienced people.
The outside is very impressive and well made and I dont want to touch that appearance, but I have taken on as my project to modify it to a much better stability & performance.

I have done some of the basic mods and removed the wax from all coils, fixed the oscillator pulling as described early in some magazines and articles, but want to take it futher by discussing the circuits here.

First thing (seen from the antenna) that puzzles me is the mixer stage. It is a standard 6BE6 Heptode valve with externally injected oscillator signal.
Normally this tube is regulated by AGC on grid3 in the span of 0-30V and used with 100V on the 2-4-grids (to accept that wide AGC range?). Anode voltage normally from 100V and up.

In adoption of the 6CS6 it is adviced in the books to apply only 30V on the 2-4 grids!

In the 9R59DE diagram below it shows 26V on those grids. There is no AGC applied on grid3 in this valve (in order to keep the oscillator more stable by signal strength as the oscillator tank is connected to grid1 even the osc. valve is separate). There is a RF-amplifier stage before that is AGC or manually attenuated to not overload the mixer. Still the mixer stage easily gets overloaded in normal use, specially in the SSB mode.

Heathkit made a reciever called ”Mohawk” long ago and interestingly in their circuit description mentioned the 6CS6 as their choice of mixer tube as superior over the common 6BE6 in mixers without AGC. the 6CS6 is the sharp cut off counterpart of the remote cut of 6BE6.
Their reason for what they describe is quite obvious, but I’ve never seen that discussed by others later on. The 6CS6 was quite new at the time and destined to other use in TV sets.
But allmost all domestic recievers was at that time using the 6BE6 regulated with AGC and no preciding RF stage, so the interest was maybe low for unregulated mixer stages.

So my theory is that Trio copied or designed the mixer stage for a 6CS6, but put in the more aviable and probably cheaper 6BA6 in the production.

Anybody here know more about this?

I’m missing instruments for measuring this, but will try the empiric way.
If you search ”EH90” here on the forum or follow the link, our member Synchrodyne from NZ has a very interesting thread of these tubes. https://www.vintage-radio.net/forum/...ht=EH90&page=2

(The 6CS6/EH90 is still plentyful and cheap due to stocks from TV-use)

The first picture is from the 9R59DE(DS) diagram
No.2, 3 & 4 are selected from the Heathkit Mohawk manual
No.5 are the original specs on the 6CS6/EH90, copied from Radiomuseum
Attached Thumbnails
Click image for larger version

Name:	A3B61178-358B-4C5E-9A09-4F8547E44DE1.jpeg
Views:	899
Size:	58.6 KB
ID:	192888   Click image for larger version

Name:	4293FEDE-C522-4231-B39B-5DC3A4217A18.jpg
Views:	281
Size:	123.7 KB
ID:	192889   Click image for larger version

Name:	F2A587AE-E966-4F48-ACC7-F05E871B203C.jpg
Views:	321
Size:	64.2 KB
ID:	192891   Click image for larger version

Name:	B3C01791-5519-4886-A4DA-0756C709CB27.jpg
Views:	190
Size:	95.9 KB
ID:	192892   Click image for larger version

Name:	020C96A8-E1C2-49FE-8ECF-28E5218D69F9.jpg
Views:	114
Size:	120.4 KB
ID:	192893  

JohanBee is offline  
Old 1st Nov 2019, 6:42 pm   #2
JohanBee
Triode
 
Join Date: Apr 2014
Location: Stockholm, Sweden
Posts: 17
Default Re: 6BE6/EK90 vs 6CS6/EH90 as frequency mixer in Trio 9R59DE ?

Edit above post: Picture No.1 is not the right one.
Tried to edit the post, but not working.
Should instead be the picture here below:
Attached Thumbnails
Click image for larger version

Name:	B994B678-F0AC-48E0-895C-9E1DA5CEBA71.jpg
Views:	318
Size:	125.3 KB
ID:	192897  
JohanBee is offline  
Old 2nd Nov 2019, 11:25 am   #3
JohanBee
Triode
 
Join Date: Apr 2014
Location: Stockholm, Sweden
Posts: 17
Default Re: 6BE6/EK90 vs 6CS6/EH90 as frequency mixer in Trio 9R59DE ?

Useful links to investigate on this matter.

Diagram of the Trio 9r59DS(DE): http://www.rigpix.com/kenwood/9r59ds_schematic.gif

Service Manual Heathkit Mohawk RX-1: https://www.manualslib.com/manual/10...-Mohawk%20RX-1

Datasheet 6CS6: https://frank.pocnet.net/sheets/191/6/6CS6.pdf
Datasheet EH90: https://frank.pocnet.net/sheets/030/e/EH90.pdf

Datasheet 6BE6: https://frank.pocnet.net/sheets/093/6/6BE6.pdf
Datasheet EK90: https://frank.pocnet.net/sheets/062/e/EK90.pdf
JohanBee is offline  
Old 2nd Nov 2019, 5:48 pm   #4
JohanBee
Triode
 
Join Date: Apr 2014
Location: Stockholm, Sweden
Posts: 17
Default Re: 6BE6/EK90 vs 6CS6/EH90 as frequency mixer in Trio 9R59DE ?

I just have read and feel I must also give a link to Phil’s very good but now closed thread for those interested in the 9R59DE(DS) here.

https://www.vintage-radio.net/forum/...d.php?t=114061

My view of modifying this receiver is that it is completely OK to modify it to a better standard internally, as long as you dont disturb the front panel original layout with new holes and switches etc.
It must be ten thousends of them still out there worldwide collecting dust, so why not fix up some of them to a better standard and put to use.
It is not now and will not be a museum piece in the future.
JohanBee is offline  
Old 3rd Nov 2019, 12:35 pm   #5
turretslug
Dekatron
 
turretslug's Avatar
 
Join Date: Nov 2011
Location: Surrey, UK.
Posts: 4,385
Default Re: 6BE6/EK90 vs 6CS6/EH90 as frequency mixer in Trio 9R59DE ?

As the pin-outs for 6BE6 and 6CS6 are the same, I'd certainly be tempted to try a swap- as both anode HT and screen grid HT (describing 26V as "HT" seems a little hyberbolic!) are notably low, it's unlikely to cause a disaster if the two valves have slightly differing optimum bias requirements.

I wouldn't be surprised if other factors beyond purely technical ones came into play around the decision to use the 6BE6- what the designer was familiar with, what was available from the contracted valve supplier, what the accountants were happy with, bulk price-wise.... It was taken as read that domestic-grade receivers would pretty much have to have a frequency-changer with a variable-mu characteristic relative to the signal grid, so these were probably made in much larger numbers than ones made with sharp-cut off behaviour. More sophisticated receivers with a greater number of signal (RF/IF) frequency stages would have sufficient AGC range with these valves alone to be able to leave the frequency-changer out of the control loop, improving frequency stability as you say. Yet it seems that even with this type of receiver, the 6BE6 was favoured over the 6CS6 when presumably the price of this class of receiver would swallow any likely price difference for the sake of performance improvement.
turretslug is offline  
Old 3rd Nov 2019, 6:23 pm   #6
turretslug
Dekatron
 
turretslug's Avatar
 
Join Date: Nov 2011
Location: Surrey, UK.
Posts: 4,385
Default Re: 6BE6/EK90 vs 6CS6/EH90 as frequency mixer in Trio 9R59DE ?

Quote:
Originally Posted by JohanBee View Post
In adoption of the 6CS6 it is adviced in the books to apply only 30V on the 2-4 grids!
I see what you mean- but I'm somewhat sceptical that this represents an absolute or limiting value for circuit operation. Whilst The Valve Museum (http://www.r-type.org/exhib/aaa1272.htm) quotes 30V as "Absolute Maximum Operating Condition", it has a link on the same page to a Philips data sheet that quotes maximum screen grid operating voltage as 100V with maximum supply voltage (i.e. to feed resistor/potentiometer) as 300V and maximum cold surge voltage for both anode and screen grid as 550V- the former two limits suggesting that 30V maximum is misleading and the latter suggesting that the EH90 is no more limited by cold physical clearances than other similar size miniatures. Whilst The Valve Museum is a useful resource, some of its assertions on individual valves I find to be contentious, even bordering on misleading, I suspect that manufacturers' specific circuit suggestions or recommendations get interpreted rather dogmatically and inflexibly (by no means the only internet source that does this....) and that 30V represented an optimum condition for its marketed role as TV FM demodulator, as opposed to a never-exceed value for all situations
turretslug is offline  
Old 3rd Nov 2019, 6:36 pm   #7
G6Tanuki
Dekatron
 
G6Tanuki's Avatar
 
Join Date: Apr 2012
Location: Wiltshire, UK.
Posts: 13,953
Default Re: 6BE6/EK90 vs 6CS6/EH90 as frequency mixer in Trio 9R59DE ?

As owner of one of the Trios in question, I'm following this thread with intrigue.

I wonder if the designers maybe dialed-down the mixer screen-grid voltage in order to get a lower noise-figure at the expense of some gain?

In a 'simple' broadcast-receiver with no RF-stage and only one IF-amp you generally needed the mixer to deliver some gain - and on medium/long-wave signals were strong - local interference generally set the noise_floor rather than internally-generated valve-noise, so you could bias the mixer to give conversion-gain at the expense of lowest-possible-noise.

On SW - particularly above 15MHz - valve-noise became more of a problem. So tweak the mixer for low-noise operation at the expense of conversion-gain, because you had a RF-amp and two IF-stages to provide gain-where-needed.


To me, the strangest thing about the Trio 9R59 design is why they used a crazy capacitive-potential-divider attenuator network in the feed from the last IFT to the product-detector. There's only a 3pF capacitor from the IFT to the prod-det's input! I can only think that this was an attempt to 'equalise' the audio-output when switching between the lossy diode-detector used for AM and the potentially-gainy SSB product-detector.
G6Tanuki is offline  
Old 3rd Nov 2019, 10:01 pm   #8
G8HQP Dave
Rest in Peace
 
Join Date: Sep 2008
Location: Solihull, West Midlands, UK.
Posts: 4,872
Default Re: 6BE6/EK90 vs 6CS6/EH90 as frequency mixer in Trio 9R59DE ?

It may have been an attempt to keep BFO voltage away from the AGC circuits.
G8HQP Dave is offline  
Old 3rd Nov 2019, 11:57 pm   #9
Synchrodyne
Nonode
 
Join Date: Jan 2009
Location: Papamoa Beach, Bay of Plenty, New Zealand
Posts: 2,943
Default Re: 6BE6/EK90 vs 6CS6/EH90 as frequency mixer in Trio 9R59DE ?

As turretslug has said, the 30 volts or thereabouts g2,g4 voltage often quoted for the 6CS6 (and 6BY6) is something of a red herring. It is applicable when they are used in noise-gated sync separator service, their original primary mission, but the same was true for the 6BE6 when used in the same service:

Click image for larger version

Name:	6BE6 Sync Separator.jpg
Views:	191
Size:	60.3 KB
ID:	193074

The 6BE6 was relatively noisy, much more so than say the heptode section of the ECH81. The 6BA7 was quieter than the 6BE6, and comparable to the ECH81. How the 6CS6 and 6BY6 compared, I don’t know. Nonetheless, 6BE6 noise may have been less of an issue when it was preceded by an RF amplifier. Marconi used the 6BE6 (X77, X727) as the first mixer in its top-of-the-line point-to-point receivers of the 1950s (HR91, HR92,93), although it was preceded by two RF amplifiers (Z77 and W77).

The use of sharp cutoff valves for mixers, usually without AGC bias, was not too unusual. For example, GE used the 6AU6 as an AM mixer in some of its consumer receivers, in both self-excited and separately excited forms. Use of a remote cutoff valve without AGC was also not uncommon, as was reduced or no AGC on the higher HF bands. In that context, a 6BE6 without AGC was not unreasonable for an HF-oriented receiver.

Perhaps a 6BE6 mixer with low g2,g4 voltage might have a shorter grid base than with normal screen grid voltage. If so, whether its was beneficial I don’t know, but perhaps Trio thought so.


Cheers,
Synchrodyne is offline  
Old 4th Nov 2019, 10:57 am   #10
turretslug
Dekatron
 
turretslug's Avatar
 
Join Date: Nov 2011
Location: Surrey, UK.
Posts: 4,385
Default Re: 6BE6/EK90 vs 6CS6/EH90 as frequency mixer in Trio 9R59DE ?

I gather (though this is only anecdotal) that the 6BE6 was essentially a scaled-down 6SA7 structure for B7g presentation, possibly with a few minor tweaks, and thus effectively represents mid '30s mixer philosophy where VHF usage wasn't really on the spec. sheet and even usage above 18MHz broadcast would have been relatively unusual. If so, ultimate noise performance wouldn't have been seen as essential. Valves such as the ECH81 and certainly the 6BA7 were post-war creatures, so lower-noise performance beyond 30MHz would have been a more significant requirement (the earlier ECH41/42 could be seen as a pioneer here), though the use of the multiplicative, multigrid frequency-changing technique generally was fairly short-lived at VHF.

It would be interesting to know how the 6CS6 compares with the 6BE6 as receiver mixer from several different aspects- signal-handling, conversion gain, noise and so on.
turretslug is offline  
Old 4th Nov 2019, 3:49 pm   #11
JohanBee
Triode
 
Join Date: Apr 2014
Location: Stockholm, Sweden
Posts: 17
Default Re: 6BE6/EK90 vs 6CS6/EH90 as frequency mixer in Trio 9R59DE ?

Thanks for your shared interest in this.
As most SW-receivers with a RF-stage many times have a need for reducing the gain before the mixer, due to overload or x-modulation, the noise in the 1:st mixer is essential. Specially when used at frequencies up to 30 MHz.

So I still beleive the Heat people when they, in the ”Introduction” text in page 5 of the RX-1 manual, states that they made a lot of efforts and ”nomerous tests to obtain a very quite tube type” and found the 6CS6 to be superior in this place compared to others including the 6BE6.

”1st Mixer
A 6CS6 tube which has dual grid control and sharp cut off characteristics is used as the 1st mixer. This tube, although not primarily designed for mixer service, has proved to have equal conversion transconductance and considerably less mixer noise than most others tested”


Link to the complete manual of the Heatkit Mohawk RX-1:
https://elektrotanya.com/cgi-bin/dow...eceiver_sm.pdf

From the diagram they are just feeding the 2,4-grids from B+ through a 47KOhm resistor. There is no voltage reading there in the diagram, but I’ll put a EH90 in my 9R59DE and take a reading. Cathode resistor is 470 Ohm in the RX-1 compared to 330 Ohm in the 9R59DE.

I supply the voltages to the mixer stage’s anode and grids from the OA2 stabilized 150V rail.

@ G6Tanuki: Regarding the product detector I’ve intentionally left it out here not to complicate this thread, but will of course take it up later, as there are some discrepances there too.......
JohanBee is offline  
Old 4th Nov 2019, 11:14 pm   #12
turretslug
Dekatron
 
turretslug's Avatar
 
Join Date: Nov 2011
Location: Surrey, UK.
Posts: 4,385
Default Re: 6BE6/EK90 vs 6CS6/EH90 as frequency mixer in Trio 9R59DE ?

Good luck- yes, the 9R59DE is something of a classic, if there wasn't an AR88D doing its stuff here, it's the sort of set I'd give space to. As you say, sympathetically improving one amongst thousands is hardly a heritage crime!

I'm always interested and encouraged to see valves used outside what might be regarded as marketing guidelines rather than technical limitations- being essentially made of metal and vacuum, there's often less inherent limitation and role-selection than in specific semiconductor recipes (staying conscious that lead length and clearances inherent in types of construction impose their own limitations on frequency and voltage capability, of course). The classic example is "small VHF PA" EC90 being doubled up as "AF signal amplifier" ECC82- both valve types being perfectly capable of fulfilling each other's and many other roles. There seem to have been a rash of post-war miniature multigrids described as "logic elements"- E91H, EH860, EH960 etc., obviously deriving from their gating capability but perhaps the generic term "modulator" is more helpful with multigrids generally whether specific types were marketed for frequency changing, detection, sync separation, logic gating or whatever.
turretslug is offline  
Old 5th Nov 2019, 2:06 am   #13
Synchrodyne
Nonode
 
Join Date: Jan 2009
Location: Papamoa Beach, Bay of Plenty, New Zealand
Posts: 2,943
Default Re: 6BE6/EK90 vs 6CS6/EH90 as frequency mixer in Trio 9R59DE ?

Quote:
Originally Posted by turretslug View Post
I gather (though this is only anecdotal) that the 6BE6 was essentially a scaled-down 6SA7 structure for B7g presentation, possibly with a few minor tweaks, and thus effectively represents mid '30s mixer philosophy where VHF usage wasn't really on the spec. sheet and even usage above 18MHz broadcast would have been relatively unusual.
In fact the connection with the 6SA7 was confirmed in RCA Application Note AN-121, in which it was said: “The 6BE6 is a single-ended, glass, miniature converter. It is equivalent in most characteristics to the 6SA7 and is similar in electrode arrangement.” The opening paragraph of that application note was: “The expanding requirements of modern receiver design have resulted in the development of several new tubes which may be used to advantage in both FM and standard broadcast bands. Two of these, the 6BA6 and the 6BE6, are new miniature tubes particularly suitable for the rf and converter positions of dual-purpose FM/AM receivers.”

Quote:
Originally Posted by turretslug View Post
It would be interesting to know how the 6CS6 compares with the 6BE6 as receiver mixer from several different aspects- signal-handling, conversion gain, noise and so on.
I haven’t found anything on this aspect. Given that the 6SC6 (and 6BY6) post-dated the 6BA7, I think that there is some chance that they were generally similar to it in noise characteristics, although it is also not beyond the bounds of possibility that one or both were direct derivatives of the 6BE6. Being sharp cutoff, they would probably have had somewhat lower noise than whatever remote cutoff valves they were derived from, if any. It would appear to have been Zenith’s use of the 6BE6 as a noise-gated sync separator that spurred the development of these two valves, which by itself might point to the 6BE6 as parent. But they were also slated for use as colour television subcarrier synchronous demodulators, for which, before their arrival, both the 6AS6 (dual-control version of the 6AK5) and 6BA7 were used or at least proposed. Whilst the choice of the 6AS6 may have been happenstance (in that very few dual-control pentodes were then available), the nomination of the 6BA7 over the 6BE6 suggests that relative noise (at the colour subcarrier frequency of 3.58 MHz) was an issue. And if so, then the 6CS6 and 6BY6 preferably would have been as quiet as the 6BA7. All speculative, though…

Quote:
Originally Posted by turretslug View Post
, though the use of the multiplicative, multigrid frequency-changing technique generally was fairly short-lived at VHF.
RCA seems to have miscued there. The 6BE6 was something of a square peg in a round hole for VHF and FM applications, so RCA quickly followed it with the much quieter 6SB7Y in 1946, on the octal base because at the time, the noval base hadn’t been thought of. In 1947 GE followed a different vector with its 12AT7 double-triode on the noval base as a much better solution for the FM mixer-oscillator application, and a miniature as well. RCA’s response, from a valve viewpoint, was to issue the 6BA7 in 1948, essentially a noval version of the 6SB7Y, even though it must have been obvious by then that the pentagrid was a dead-end for FM applications. Although for its own receivers, as a stopgap measure, it used the 6J6 until c.1951, when the 6X8 became available. The 6BA6 was also a bit of a push for FM, and the improved 6BJ6 was soon issued. Zenith, a major proponent of FM in the early days, soon adopted the 6BJ6/12AT7 combination for FM (and FM/AM) front ends, and then used it for a decade or more. With RCA, one could wonder “what were they thinking?”. Although RCA had had some excellent FM engineers (e.g. Crosby, Avins) one may wonder if Sarnoff’s negative attitude towards FM quashed some FM-specific developments.

Returning to the situation at interest, a point to note is that in the case of the Heathkit RX-1, the 6CS6 mixers appear to have been used with the signal on grid 1 and the oscillator on grid 3. The 6BE6 was normally used with signal on grid 3 and oscillator on grid 1, both for the self-excited and separately-excited cases.


Cheers,
Synchrodyne is offline  
Old 5th Nov 2019, 10:13 am   #14
turretslug
Dekatron
 
turretslug's Avatar
 
Join Date: Nov 2011
Location: Surrey, UK.
Posts: 4,385
Default Re: 6BE6/EK90 vs 6CS6/EH90 as frequency mixer in Trio 9R59DE ?

I assume that, with most multigrids, grid 1 is able to exert greater influence on the cathode current than grid 3 (or whichever is the other designated control grid) and results in higher conversion conductance- apparently the reason for the grid 3 link with triode grid in most triode-hexode/heptodes (other than the 6K8 whose construction effectively forces grid 1 linking). That could be Heathkit's rationale for applying signal to grid 1 here, especially when they were seeking to further optimise the usage of a "different" valve. I'm also assuming that classical US heptodes such as the 6SA7 and 6BE6 had grid 3 as a variable-pitch grid for AGC functionality and that grid 1 (the oscillator grid in the self-oscillating converter topology) had an even-pitch construction, so that even in externally-driven service, the signal was better applied to grid 3 when the valve was subject to AGC action to minimise oscillator pulling- the avoidance of the latter being a principal motive for adopting external drive.

I'd wondered why such sets as the AR88 that use externally driven 6SA7 with no AGC application continued to use grid 3 as signal grid- but I have heard it said that grid 3 characteristics can include a negative resistance component in action, improving input circuit Q over the use of grid 1. If so, it represents an interesting design quandary between the choice of grid 1 (higher conversion conductance) or grid 3 (less RF circuit damping).
turretslug is offline  
Old 5th Nov 2019, 8:28 pm   #15
turretslug
Dekatron
 
turretslug's Avatar
 
Join Date: Nov 2011
Location: Surrey, UK.
Posts: 4,385
Default Re: 6BE6/EK90 vs 6CS6/EH90 as frequency mixer in Trio 9R59DE ?

Come to think of it, applying an LO signal of several volts amplitude to a variable-pitch grid would likely produce appreciable distortion products (LO harmonics) in the anode current when there is enough of a headache keeping spurios as a result of electron stream modulation action low anyway, so only injecting the oscillator at a grid with a reasonably linear large-signal transfer characteristic wrt anode would make sense. With the 6CS6 not being a specific variable-mu type, that presumably means that there is the option of injecting the high amplitude LO at either of the controlling grids.
turretslug is offline  
Old 5th Nov 2019, 9:06 pm   #16
G6Tanuki
Dekatron
 
G6Tanuki's Avatar
 
Join Date: Apr 2012
Location: Wiltshire, UK.
Posts: 13,953
Default Re: 6BE6/EK90 vs 6CS6/EH90 as frequency mixer in Trio 9R59DE ?

Considering this, I'm wondering about the WWII Navy TCS12 receiver which used a 12SA7 as the mixer.

Crazily, it used a 12A6 beam-tetrode as the local-oscillator. Yes, a companion to the 6V6 - capable of brewing-up a few Watts of LO power. Was this done for some particular reason? It seems mad but there must have been a reason.
G6Tanuki is offline  
Old 5th Nov 2019, 10:58 pm   #17
Synchrodyne
Nonode
 
Join Date: Jan 2009
Location: Papamoa Beach, Bay of Plenty, New Zealand
Posts: 2,943
Default Re: 6BE6/EK90 vs 6CS6/EH90 as frequency mixer in Trio 9R59DE ?

As far as I know, when the 6BE6 and 6SB7Y/6BA7 were used in the separately excited mode, grid 1 was still used for oscillator injection and grid 3 for signal input. But right now I cannot find a specific valve maker reference to this, although I think I have seen it somewhere. It is not covered in the application notes, which by the way are available here (with some interesting editorial comments): http://www.one-electron.com/Archives...-AppNotes.html.

With the 6CS6 as mixer, I’d guess that Heathkit’s use of grid 1 for signal and grid 3 for oscillator was determined empirically – it probably tried it both ways around, and used what was the better of the two for its purposes.


Cheers,
Synchrodyne is offline  
Old 6th Nov 2019, 11:06 am   #18
turretslug
Dekatron
 
turretslug's Avatar
 
Join Date: Nov 2011
Location: Surrey, UK.
Posts: 4,385
Default Re: 6BE6/EK90 vs 6CS6/EH90 as frequency mixer in Trio 9R59DE ?

Thanks for that link Synchrodyne- that made for an absorbing browse, with stacks still to digest! Amongst plenty else, it refreshed and deepened my understanding on the g1/g3 playoff, also the rationale behind quite a few receivers with US-pattern heptodes going osc-low on their highest HF band in order to optimise space-charge phasing and prop up conversion conductance.
turretslug is offline  
Old 6th Nov 2019, 1:16 pm   #19
turretslug
Dekatron
 
turretslug's Avatar
 
Join Date: Nov 2011
Location: Surrey, UK.
Posts: 4,385
Default Re: 6BE6/EK90 vs 6CS6/EH90 as frequency mixer in Trio 9R59DE ?

Quote:
Originally Posted by G6Tanuki View Post
Considering this, I'm wondering about the WWII Navy TCS12 receiver which used a 12SA7 as the mixer.

Crazily, it used a 12A6 beam-tetrode as the local-oscillator. Yes, a companion to the 6V6 - capable of brewing-up a few Watts of LO power. Was this done for some particular reason? It seems mad but there must have been a reason.
Is this a case of a physically large valve but running at relatively low anode current/power? Perhaps they were aiming at relatively low temperature rise for stability. Despite being presented as a power valve it still has the typical 2W heater of signal valves. This set took care over frequency stability with its wide-spaced oscillator gang section, and Collins give the impression of knowing what they were at, so they must have done what they did for good reason- I must dig the circuit out. It might be that being an audio-orientated valve with relatively high grid bias that it was able to produce a good oscillation amplitude but with relatively low harmonics.
turretslug is offline  
Old 7th Nov 2019, 12:13 pm   #20
JohanBee
Triode
 
Join Date: Apr 2014
Location: Stockholm, Sweden
Posts: 17
Default Re: 6BE6/EK90 vs 6CS6/EH90 as frequency mixer in Trio 9R59DE ?

Thanks Synchrodyne for the link to the interesting RCA papers, specially about the 6AS7.
They say there should be an osc. voltage enough to give a small grid current in grid1 to operate ok in case of separate fed osc. injection. Also in the diagram example they show, the osc. signal is brought out from the cathode tap in the osc coil. Will preserve the Q and stability better in the coil, if enough voltage there.
In the 9R59DE the injection is fed via a 5pF cap from the top of the tank circuit and loads down the osc. considerably at higher frequencies. Trio has increased the grid1 leak-resistor to 100K from the recommended 20K, probably to easen the osc. load.......

I’ve now run a new Philips valve marked EH90/6CS6 in my 9R59DE and the receiver is working OK as before on the 3 lower bands. I’ve measured roughly the same voltages and currents readings as with the 6BE6.
Not any 26 V ever on grids 2-4!

@ B+ 146V from the 0A2 stabilizer valve.

6BE6: V-G24 =35V with original 68KOhm resistor to B+, with 47KOhm =42V, Ik=2mA
6CS6: V-g24. =49V with 47KOhm resistor, Ik=2,7mA, Ia=0,6mA

The oscillator voltage in my 9R59DE is still very weak on the top band up to 30MHz. I think this is the source of the problem more than, or together with the operating parameters of the mixer valve. (from the beginning with the thick wax lump it was not oscillating at all)
I have (on RCA’s designer guides suggestion) changed the osc. valve socket from the brown ”micanol” type to a ceramic with a real shield skirt, to decrease the termal drift of the oscillator. Also changed the osc. valve to a cooler running 12AT7/ECC81.

Maybe I’ll have to take out the osc. coil to clean it from the residual wax between the windings, or clean the coil former and rewind it with new wire.
Or maybe first make a new provisional outboard osc. coil, to check the feedback windings No’s is working OK. The diagram shows taps on the osc coils, but IRL it is separate feedback windings on the formers.
On L112 there is a strange second feedback winding I dont understand why it’s there for, fed from the osc. anode via. a 0,01u capacitor and a 150 Ohm resistor from B+ so there are 2 feedback windings on that coil!
That winding is not disconnected when changing bands.

The more I study the R59 I feel the design has so many more flaws than discovered at the first glance.
But it has a very nice appreance on the outside and front panel layout, a good selectivity in the IF (peaking at 456KHz) and easy to work with, so I never give up to make it work properly.

BTW, a good tip is to change the power gustling 6AQ5/EL90 to an EL95. Can be swapped with no other change of components and easen the burden on the mains transformer. Also suits better powerwise to the small output transformer.
The 3pcs 6BA6 can also be swapped to 6BJ6, to save another 450mA heater current.......(for a lamp to the S-meter).....?
JohanBee is offline  
Closed Thread

Thread Tools



All times are GMT +1. The time now is 10:17 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.