Re: The "Sussex" Homebrew Valve Tester.
 

Hello Peter,

The transformers were wound from scratch. I have plenty of laminations and cores, but only one size, about 50 VA.
I must admit I did not check Farnell, thanks for that info.
It's a pity that the heater winding is in the middle, but why don't you do what I did and wind two separate windings of 12v AC and then, with the addition of an extra switch, a relay and an inexpensive DC to DC convertor from China, you can have any voltage you want just by turning a 10 turn pot on the front panel. Also less noise.

The DPM's input impedance can be measured by using you meter set on ohms, but I double check by looking at the resistances value printed on the resistor itself.

Yes, my Tester is coming along nicely now.

I have a meter to measure anode voltage and current, so I can also measure other valve parameters as suggested by Peter.

It's making PCB's that is the problem.

Karl

Re: The "Sussex" Homebrew Valve Tester.
 

Hello!

I am really interested in your .fpd files for "sussex" tube tester.

Can you share them, please?

Thanks, best regards.

Fast

Re: The "Sussex" Homebrew Valve Tester.
 

Hello,
I received two DPM's yesterday - 200mV DC and 200mV AC. I tested them but I'm getting very strange results.

For the power supply I'm using 6.3V heater winding from a tube amp transformer and the test voltage is another 6.3V from the same amp. I made a voltage divider from a 150K resistor and 5K potentiometer in series. To test the DC meter I added a bridge rectifier and a 470 uF cap before the divider.

I don't have an accurate DMM at home so I'm using an analog Unigor multimeter. I will give the Unigor readings in brackets:

1. 200 mV DC DPM:

50 mV (44 mV)
100 mV (88 mV)
150 mV (133 mV)
200 mV (176 mV)

Then I replaced the Unigor with another cheap DMM and it gave almost the same readings as Unigor (1-2 mA difference only) at the same readings of the DPM.

2. 200 mV AC DPM (50Hz):

50 mV (36mV)
100 mV (72 mV)
150 mV (110 mV)
180 mV (145 mV)

3. 200 mV AC DPM (1kHz sinewave from PC headphone output):

47 mV (33.8 mV)
100 mV (71 mV)
173 mV (141 mV)
200 mV (199 mV)

It seems that the AC DPM is calibrated at 200 mV but gives very unlinear results.
Am I doing something wrong or these meters are unusable? I will repeat the measurements with battery power supply to see if it will help.

Peter

Re: The "Sussex" Homebrew Valve Tester.
 

I've found my mistake - I have supplied the meters with 5.4V AC instead of 6.4V. When I repeated the test with batteries I've got good results - the difference with the DC meter was only after the dot.
With the AC meter the maximum error was 1-2mV compared to Unigor readings.

Re: The "Sussex" Homebrew Valve Tester.
 

Hi Peter, read the notes carefully about the supply and input terminals on these meters as some have different rules about commoning link between supply and measuring circuit.

Ed

Re: The "Sussex" Homebrew Valve Tester.
 

Hi Ed,
Yes, I've read that part - I am using different sources for the supply voltage and the test voltage. But I used by mistake a heater winding for GZ34 which gives about 5V instead of 6.3V.

I've noticed also that if the supply voltage is low (8.4V) the AC meter start giving wrong results. With 12V battery it was working properly. So I plan to make the meters' supply windings to give about 7.5-8V AC.

I also checked the input shunt resistors - the 200mV AC meter has one 10K 1% resistor with leads and the the 200mV DC meter has three 698K SMD resistors in series (almost 2.1 MOhm).

Peter

Re: The "Sussex" Homebrew Valve Tester.
 

Hello Peter, Ed and All,

Good work there with your testing of the meters!
Perhaps what was said earlier in this thread about the all the AC 200mV meters being lacking in accuracy are put to rest...
What and where did you buy your meters? Perhaps other forum members may be interested in the info.

But I am still sticking to my DC meters and will use my little AC differential amplifier and RMS to DC convertor - it gives me gain as well as better resolution.
Your dead on with the meter windings output - all seven of my meter windings are wound for 9.0 AC, which gives me about 13V to 14V DC.

I know I keep on about my 200mV DC meter method, but look - the 1KHz AC signal from the 10 ohm sense resistor is capacitor coupled and amplified by my X10 gain diff.amp to give me signal AC mV X 10, this is then converted to RMS DC to give me exactly DC mV X 10. Now, instead of dividing down the X10 signal, I feed this signal as it is, directly to the 200mV DC meter, which of course will now read 10 times higher - but now I merely move the decimal point on the meter to the next left position(DP2) so that the meter reads the value correctly, but now with a resolution of 0.01mV i.e. a valve with a GM of say, 5mA/V, will read 5.00mV on my meter.

The other advantage of course, is that the coupling from the 10ohm anode sense resistor is now AC, with the result that the meter circuit itself is not at high anode voltages and will be in a much less dangerous position if something goes amiss with the anode voltage regulator.
I think, if I remember correctly, that a few forum members had the anode GM meter destroyed when a fault occured on the anode regulator FET.

Ed, was it you who put a 1N4007 diode across the 10 ohm sense resistor to save the DPM if something went amiss?

I am now wiring up the various modules, but it is a long hard slog…

Ed is right about the supply requirements of these DPM's...some share a common 0V,
and others need to be kept separate, but I have noticed on many DPM's, although supplied with a 4 wire lead and the suppliers state that the supply and signal must be kept separate, that in fact the power 0V and the signal 0V are actually joined on the meters PCB!

Peter, how is your build going?
Have you wound that transformer yet?

Just one more thing...As I am also monitoring the Anode current on one of my many meters, when I am actually testing a valve, what if I change, say, the negative grid voltage by exactly 100mV down, and note the anode current, then change the negative grid voltage exactly 100mV up, and again note the anode current, then, the change in anode current from the down 100mV, and the up 100mV, and this value divide by two, will also give the GM of the valve!!

Another way to double check the GM readings?

Ed, Peter, is this Right, or have I lost the plot?

Karl

Re: The "Sussex" Homebrew Valve Tester.
 

Peter Ed and All,

That last bit, the change in mA is first divided by 2 and the multipied by 10...I think...

Karl

Re: The "Sussex" Homebrew Valve Tester.
 

Karl,

I have sent you a PM re source of my meters, as I am not sure if I can say where on the open forum.
Keith

Re: The "Sussex" Homebrew Valve Tester.
 

Hi Karl,
I wanted to buy the meters from Giorgio (Asia Engineer) but he didn't had the AC meter so I ordered them from this store: http://stores.ebay.com/haruyrkb
They look the same as those of Asia Engineer are currently selling and have green terminals with screws. If there is interest I can post pictures of them. Does yours have the same input resistances?

I checked for continuity between the supply and input terminals - they don't have common 0V. On the AC meter I measured resistance above 1M and on the DC meter it's out of my DMM range (2MOhm).

I finished winding the transformer but have not assembled and checked it yet. Also have to buy some capacitors and to get an enclosure (a friend promised to give me one from some vintage device).

It will be interesting if you post the schematics of your AC differential amplifier and RMS to DC convertor and also the oscillator you are going to use.

About the DC-DC converter for the heater supply - aren't these quite noisy?

You are right about the calculation of Gm - it is similar to those of Ra. If the change of Vg is 100mV (or 1V for example) and the Ia is in mA, you have to divide the change of Ia by 100 or 1000. In your case - by 200. The important thing is if we change one of the voltages (Vg or Va) we must keep the other the same and record the change of Ia.

Now I'm trying to draw the schematic and the board with Eagle PCB but the shareware version has a limitation of maximum size - 80x100mm. So I have to split the main board in two parts. I will check if is possible to fit the relay board in the part with the relay and bias supply.

Peter

Re: The "Sussex" Homebrew Valve Tester.
 

Hello squegging,

Thanks for the info...
How are you getting on with the tester?

Karl

Re: The "Sussex" Homebrew Valve Tester.
 

Hi Karl,

Still collecting components at the moment. Have one of Ed's transformers plus DPM's from ithingsonlineshop on eBay. Would be interested how you use the 8038 oscillator, as that seems preferable for amplitude stability. Also plan to add 2v dc heater supply as I have an old radio circa 1935 that my grandfather made. He had the first "wireless" in Lostwithiel, Cornwall as far as I know. I'm not using regular pcbs, but will build on tagstrip / veroboard. Must get round to ordering those switches etc....

Regards
Keith

Re: The "Sussex" Homebrew Valve Tester.
 

Hi Ed and Martin,

I am not sure how to send PM. I am interested to build this tester and need to buy the Transformer and PCB if still available.
Please advise.


Bernard

Re: The "Sussex" Homebrew Valve Tester.
 

Hi Bernard, as a new member your PMs are limited. Easiest way is to click on the member's name and use the send e-mail option.


ed

Re: The "Sussex" Homebrew Valve Tester.
 

Hi Andy,

It appears that Martin has no longer have the first batch PCB.
Is there any plan yet to commission this new one?

Bernard

Re: The "Sussex" Homebrew Valve Tester.
 

Hello Peter, Keith and all,

I have attached a rough drawing of my heater control wiring (some of it anyway)
I have made many measurements and tests and the results are very good indeed.

Peter, this DC/DC convertor (Buck) allows me to set any DC voltage between 1.5 V DC to about 28 V DC - I use the heater control switch (4P 5W) to select fixed outputs of 4V DC, 5V DC, 6.3V DC, Variable 1.5V DC to 28V DC, and lastly, External input.

The lower voltages (up to about 12V DC) are preset by the 3 X 10K presets and can supply up to about 2 Amps. These preset heater voltages can be any value - indeed, if you use more poles on the switch, say 8 poles, then you can have 6 different heater settings, and the variable and the External input.
I switch two 12V AC windings from one of my transformers, either in parallel for the lower, high current heater voltages, and then switch the two windings in series for the higher lower current heaters...this keeps the power, and heat loss,to a minimum across
the Buck Convertor.
Of course, one has to remove the onboard 10K preset on the DC to DC convertor PCB and attach two wires to the heater control switch as shown in my diagram...

But another way is to use 3, or more as required, of the PCB convertors and then one doesn't have to do any work on the PCB's!
The Buck DC convertors are very cheap!

Now, ripple...or lack of it - with a pair of 6L6's in my test setup, therewas max. 8mV ripple at 60KHz, and about 1 or 2 mV 100Hz.

Another not so obvious benefit, is the fact that one doesn't have to bother about wether the AC mains supply is 220, 230, 240 or whatever - if 6.30 V DC is set, then it will stay there, and with much reduced hum and noise pick-up.
Yes, there is more wiring involved, but the benefits outweight the cons.

With the Variable control selected, as I said earlier, the two 12V AC windings are now in
series, and heater voltages up to 28 V DC can bet set by the front panel 10K 10 turn pot.

In fact, I am wiring this part up as I speak - I cannot wait to finish the Valve Tester!!!

How are other builds coming on?

Re: The "Sussex" Homebrew Valve Tester.
 

Hi Karl and all other Sussex builders,

I was on a vaccation at the seaside but now I'm home and ready to continue the build of the tester.

Thanks for the info about those DC-DC converters. There are several different types - which one do you recommend? Can we use another step-up converter for voltages from 35-60V? Now I have to dissasemble the transformer again and modify the heater's windings!

I started drawing new PCB's with the later changes in the schematic. When I added screw terminals it get complicated so I have to optimize them a little.

I still wonder how to make the oscillator - that ICL8038 chip is an obsolete part - is there a similar replacement part?

I can post the schematics that I have drawn for the PCB's so we can make additional changes or remove errors. Then I will correct the boards and will post them for final checking and use from the forum members.

Karl, with the AC-DC converter can't we use only one 200mV DC panel meter for Gm and Ia (and Ig during Gas test)? We can also connect the meter across the 10 Ohm resistor when testing preamp tubes to increase the precision with low anode currents. Can we damage the meter if we forget the switch in that position and the voltage is 1.8 - 2V with some power tube?

Best regards,
Peter

Re: The "Sussex" Homebrew Valve Tester.
 

Hello Peter, and all builders of the 'Sussex'

The DC-DC buck convertors I am using at the moment are from eBay site 'polida2008' and from this site I get most my semiconductors e.g. diodes, rectifiers, transistors and I.C.'s. There are very professional and very reliable.
The DC-DC convertors are about £1.30 including postage, unbelievable!

About higher heater voltages - yes, I also need to test valves from AC/DC radio's, and television valves, but I was going to use the 'External Heater Supply' position on the heater control switch to achieve this.
But your suggestion of internal facilities for higher heater voltages has got me thinking...

Of course, the ‘Variable Heater Voltage’ position on my present set-up does allow me control of heater voltages from about 1.5V to about 28 Volts DC at currents of 300 to 400mA, so this is a good start.
But I agree with you, it would be nice to have higher heater voltages to cater for all types of valves, within the tester itself.
The ICL3038 Oscillator chip is still available for about £1.50 for quantities of 10off, I will post my ICL8038 Oscillator circuit as soon as I can - it will be hand drawn as all my notes on the build details are in a big fat A4 folder, with many, many pages.

I think we can use the original circuit and still use my heater control circuit - I will post the whole of this heater control circuit, and perhaps you and others may find it useful, or suggest useful changes.

Your idea of using the 10 ohm (anode current sense resistor) resistor to enhance the GM readings, instead of the 1 ohm GM resistor for low current valves is good - however, the switching may be difficult to arrange.
Again, with my GM amplifier circuit, this may not be necessary as we have a gain of 10 compared to the original circuit (and DC heaters) and this should give good results, as my tests have indicated. The X10 Amplifier chip can also be obtained from ‘polida2008’ for a couple of pounds – AD620.

The more I look at the overall design, it would appear that there are numerous ways of achieving the same end result, but I am confident, with the tests I have carried out, that my tester will give me what I want.

The biggest problem I have at the moment is that I will be, in a day or two, going into hospital to have my left eye operated on, but after the op, and a few days recovering, I will be back on the job.

I have only just started to wire up all the modules in my tester, so I can still make changes to the overall design if I need too...

Now, what valve sockets do we need to make sure that most valves can be tested?
Peter, what sockets are you fitting?

What parts of the wiring should be screened cable?
How on earth am I going to annotate the front panel?

Perhaps other builders who have already built the tester can help on these and other issues?

Peter, I have been thinking very seriously about PCB's - I may, just may, splash out on a basic kit to make my own PCB's - this seems to be a stumbling block on big projects like this. Can you and other forum members give advice on this?
I have very good PCB software (EASY PC) so with a UV exposure unit, a laser printer, some Ferric Chloride, developer etc. perhaps I can make my own PCB's...

Peter, have you re-wound the heater transformer for two 12V windings?

[ And then the phone rang…Royal Bournemouth Hospital…we have a cancellation, can you come in tomorrow for your operation on your left eye – panic, yes I will come in..]

That was 2 weeks ago – my left eye is unusable at the moment, but is healing nicely…

I have attached a couple of very rough drawings of my DC heater circuit and the ICL8038CCPL 1Khz oscillator circuit. The 8038 can still be bought for about £2.00 each from the same eBayer as the AD620.
I will send in the next couple of days, a drawing of my GM measurement circuit.
I have been out of it for a couple of weeks now, what progress on other builds?
Karl

Re: The "Sussex" Homebrew Valve Tester.
 

Hi Karl, on the question of pcb's; it may be possible to obtain the masters or Gerber files of the original units if it can be determined where they were made.
Copyright issues would need to be considered or permission sought. This would probably be a lot easier than a self etch route.
I use Easy PC professionally and it is will produce much better artwork than could ever be self etched.
If you do a new design then it should be no problem getting a PC house to quote you for a few sets, assuming there is interest here from members. Prices are relatively low these days if you look in the likes of Elektor.


Ed

Re: The "Sussex" Homebrew Valve Tester.
 

Hello Karl, Ed and the other builders,

Karl I was wondering where were you and I'm glad to hear that you are well and recovering from the operation. Also thank you for the schematic.

I will check that polida2008 seller but I haven't modified the transformer yet.

I am ready with the updated version of the main PCB - even tried to transfer the image on a board from glossy paper (from a brochure). After several unsuccessful tries it's not so bad now - needs only some minor corrections that can be done with paint. Next week I will buy some Ferric Chloride and etch it.

The updated main board is based on the original design by Mike Rowe with the changes made by the group and published by Les Carpenter. I have used Eagle PCB software (with the Freeware license for "non-profit" and evaluation purposes) so if it is not considered as a violation of copyrights I can post the images and the schematic.

Karl, with your permition I can make two versions of the second PCB - one with the original oscillator and other with your version. The only problem I see is that your oscillator needs additional power supply.

I build tube guitar amps in my free time and know that some guys are using water decals for the faceplates but haven't tried that yet. Other option is laser engraving but it may be expensive.

In addition to the standard seven, nine and eight pin sockets I have some vintage ones that I can also use (F8, loctal and P-type). I'm still looking for other types.

I think it's neccessary to use screened cable in the grid wiring only (on the Les' schematic labeled as "Control"). Are you going to use ferrite beads on the wires near the sockets?

My friend will give me the box next week (it needs cutting an opening at the back side) so I will continue with my build.

Peter