|
Homebrew Equipment A place to show, design and discuss the weird and wonderful electronic creations from the hands of individual members. |
|
Thread Tools |
9th Feb 2014, 9:34 am | #1041 |
Rest in Peace
Join Date: Mar 2011
Location: Basildon, Essex, UK.
Posts: 4,100
|
Re: The "Sussex" Homebrew Valve Tester.
The relays arrived Saturday afternoon and were fitted.
I did a full live test of the system and found 2 major problems, not found previously when I was testing each module separately. 1. I had mistakenly joined the ground for the HT with the negative supply rail of the bias circuit. I got through a few fuses before I found the problem, I am glad I fitted fuses in the bias supply!. This could be a useful warning to any new builders of this project. DONT MIX THE 2 -VE SUPPLIES. 2. I had a break in the copper track of the anode current limiting circuit. This had happened by accident when I first made the circuit last year in the summer. At that time I had bridged over the small gap with solder, but more recently I changed R6 from 100R to the correct value 120R. In doing so I inadvertently removed the solder bridge resulting in the base of TR3 and the hot end of R6 being disconnected from the rest of the circuit. There was probably therefore no proper current limiting. This was not evident when testing triode and pentode valves but became very evident when I tested a rectifier valve, and another fuse (HT) went pop and my AVO which was monitoring mains current went off scale. Both issues now fixed and I declare the project working well. I have a bit of learning and familiarising myself with valve testing and few questions and queries on the tests I have carried out so far on old valves, I will read through the notes in the manual on testing and then collate any questions to ask for some advice, Cheers Mike Last edited by crackle; 9th Feb 2014 at 9:45 am. |
10th Feb 2014, 10:10 am | #1042 |
Rest in Peace
Join Date: Mar 2011
Location: Basildon, Essex, UK.
Posts: 4,100
|
Re: The "Sussex" Homebrew Valve Tester.
Hi
I have a few questions about using my Sussex and would appreciate some help. 1/ When testing EL34 valves I do not get anything like a gm of 11, it is more like 2 to 3.5 mA/V. The anode current rises and the gm rises to about 5 but then the gm falls back. Is this something to do with needing a grid stopper that I have read a couple of pages back. Is the grid stopper left in circuit always, what is the optimum value, will it effect the "gas" readings. 2/ Testing a EL84 valve has some strange effects on the trace of the grid signal, see 2 photos below, 1 before valve has fully got "hot" and one showing reduced amplitude and multiple waveforms when valve has been operating for a time. 2/ I am not sure when to perform a gas test, is it whilst the valve under test is operating and the function switch is set to "test"? Thanks Mike p.s. When testing some EL33 valves I was getting quite believable results for Anode current (about 40mA) and GM (from 3 to 4.4 mA/V) |
10th Feb 2014, 11:07 am | #1043 | |
Rest in Peace
Join Date: Mar 2011
Location: Basildon, Essex, UK.
Posts: 4,100
|
Re: The "Sussex" Homebrew Valve Tester.
Quote:
Thinking about this a bit more it is not going to be simple to add a grid stopper to pin 5 of the valve holder as pin 5 may not always be used as the grid, (or will it). I have tried adding a 4.7k resistor in the patch lead to pin 5 and it seems to make no difference. The GM value climbs to about 5 then falls back to 2, even on a newish valve, a "Groove Tube". Please can anyone offer some help and advice. Thanks Mike |
|
10th Feb 2014, 11:45 am | #1044 |
Moderator
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,797
|
Re: The "Sussex" Homebrew Valve Tester.
You need a scope in there to measure real voltages and currents at various places to see what's going on. Those digital panel meters will not handle AC components well if there is any instability.
David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done |
10th Feb 2014, 12:27 pm | #1045 |
Rest in Peace
Join Date: Mar 2011
Location: Basildon, Essex, UK.
Posts: 4,100
|
Re: The "Sussex" Homebrew Valve Tester.
Hi
Just to be clear, did you mean a scope across R12 to measure the AC voltage and then work out the current through the 10R resistor. Thanks Mike |
10th Feb 2014, 12:31 pm | #1046 | |
Nonode
Join Date: Aug 2010
Location: Resolfen, Wales; and Bristol, England
Posts: 2,588
|
Re: The "Sussex" Homebrew Valve Tester.
Quote:
On my tester, I simply added a grid stopper resistor soldered directly to the pin 5 of the octal valveholder and connected the wires to that. For me, that is enough since any octal valve I have looked at uses pin 5 as the grid. The resistor is something like 150k (I don't remember exactly but it's not critical, just needs to be somewhere in that region) Just to check, when you are testing EL34s, put your hand around the valve when it's being tested and see if the values on the gm and Ia meters change - in my case the capacitance from my hand was enough to settle the device down and provide a beleivable reading! Hope this helps, Richard |
|
10th Feb 2014, 3:28 pm | #1047 |
Rest in Peace
Join Date: Mar 2011
Location: Basildon, Essex, UK.
Posts: 4,100
|
Re: The "Sussex" Homebrew Valve Tester.
I have taken a couple of photos of scope images for the HT rail and the signal across R12
The first shows a trace with what appears to be multiple signals across R12. The second photo shows the HT AC "ripple" which builds up to 12v PTP when switched to "test" Any ideas if this is normal or not. Thanks Mike |
10th Feb 2014, 3:49 pm | #1048 |
Moderator
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,797
|
Re: The "Sussex" Homebrew Valve Tester.
Can't see where the CHA knob is set to, or whether the variable is in the cal position. R12 is at anode potential so you'll only gat a trace on the screen with AC coupling. Are you using a x10 probe, a x1 probe or what?
There seems to be a kHz sine there from the grid voltage modulation. but the meters sense across the resistor. Scoping one end will see the Ia modulation times ten ohms superimposed on the HT ripple. Anyway, there doesn't seem to be any low RF oscillation. I'm a bit wary of unknown panel meters on AC ranges. David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done |
10th Feb 2014, 9:12 pm | #1049 |
Rest in Peace
Join Date: Mar 2011
Location: Basildon, Essex, UK.
Posts: 4,100
|
Re: The "Sussex" Homebrew Valve Tester.
Hi David
Picture 1 ChA is set at 2 mV/div, timebase is set at 0.2mS Picture 2 ChA is set at 0.2 V/div, timebase is set at 5mS What is that sawtooth waveform? Is it generated by the primary voltage regulation or the current limiting circuit? Thanks Mike |
10th Feb 2014, 10:14 pm | #1050 |
Moderator
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,797
|
Re: The "Sussex" Homebrew Valve Tester.
Hi, Mike
So photo 1 shows about 6mV p/p sinewave at about 1kHz. Now I think the Gm test grid drive is supposed to be 1kHz, and TR6 is what's called a phase-shift oscillator (actually all oscillators are in one way or another) and should give a reasonable sinewave shaped output. But it should be 100mV RMS which means 282.8 mV peak to peak applied to the grid via C7 if the level control pot VR2 is set right. Try this one again without a valve in. It looks like the oscillator is running but the level reaching the valve under test is rather disappointing. Is something not passing the signal or is something shorting it? Photo 2 is also interesting. The power supply uses a bridge rectifier, and so long as all four diodes are happy with life, any hum should be at 100Hz. And we've got a sawtooth at 50Hz. So the first step is to see if the sawtooth is on the HT rail at the output of fuse 2. Then look for it on the output of the HT reference regulator, the collector of TR1, then follow it through to the selected tap labelled ANODE V Next comes the MOSFET source-follower. It could be current-limiter action. Check (with power off) resistance from AJ to AK and you should see 6.31 ohms. Check R3&R4. It's just steady progressive checking. Use the scope, with sawtooth AC in interesting places DVMs will be misleading. It's a nice DC coupled scope so maybe you need a 10:1 scope probe with a high enough voltage rating. Don't bother with switched x1/x10 probes. You rarely need x1, and it compromises the high frequency performance on x10... and everyone accidentally slides the little switch by accident. with 1 meg feeding the source follower gates, the current limit transistors don't have to leak much to upset things. David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done |
11th Feb 2014, 12:26 am | #1051 |
Rest in Peace
Join Date: Mar 2011
Location: Basildon, Essex, UK.
Posts: 4,100
|
Re: The "Sussex" Homebrew Valve Tester.
Hi David
Without the valve in, I have a good clean sine wave on the grid circuit. I have adjusted it to 100mV using a DMM and then checked it again with the scope which seemed to agree. The HT rail is cleen with no noticeable AC ripple the sawtooth wave form appears when the HT is switched on "test" it builds up over about 3 or 4 seconds, and it also does this without any valve in circuit. I have noticed that when a valve is in circuit the HT current also builds up slowly as if the heaters are warming up but they are already warm. Tomorrow I am going to double check my figures and the where I had the scope connected for the messy sine wave photo. I will also do some more measurements to see where else the sawtooth trace is getting to. Thanks Mike |
11th Feb 2014, 7:13 am | #1052 |
Moderator
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,797
|
Re: The "Sussex" Homebrew Valve Tester.
Does the screen voltage go up as slowly?
Those big MOSFETS can have a few thousand pF gate capacitance, and with 1 meg driving them that would still only give a few milliseconds time constant. To get such a slow rise takes either a large capacitance, a high resistance or a small current or a combination of the above. This is a nice easy clue that something is odd. It may well tie in with the sawtooth. Track the slow rise back to its source and you may find the cause of several symptoms. David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done |
11th Feb 2014, 8:35 am | #1053 |
Rest in Peace
Join Date: Mar 2011
Location: Basildon, Essex, UK.
Posts: 4,100
|
Re: The "Sussex" Homebrew Valve Tester.
I will check the screen voltage to to see. There is a 22uf cap C13 connected across the anode voltage selector on the schematic, but I mounted this on the power supply circuit module, the electrical connections should be the same.
Mike |
11th Feb 2014, 12:16 pm | #1054 |
Hexode
Join Date: Dec 2009
Location: Leeds, West Yorkshire, UK.
Posts: 250
|
Re: The "Sussex" Homebrew Valve Tester.
Hi,
I don't know if my thoughts may help? On the first Sussex I built, like you, I mounted the Mosfets off the board and used flying leads to connect them. The anode and screen voltages were all over the place and up and down like the proverbial fiddlers elbow. There were all sorts of strange oscillations. A colleague suggested that they should be mounted on the board so the gate resistors are close to the devices. Problem cured. Just a thought..... Good luck!
__________________
Howard G7AJN/M3OCL "How hard can it be?" - Jeremy Clarkson |
11th Feb 2014, 5:16 pm | #1055 |
Rest in Peace
Join Date: Mar 2011
Location: Basildon, Essex, UK.
Posts: 4,100
|
Re: The "Sussex" Homebrew Valve Tester.
Hi Howard
Thanks I will give that a try as well and relocate R3 & R4 to the gate leg of the transistor. Can anybody tell me why it needs 1.1M, made up with 2 resistors, is 1 meg not sufficient and 1.2m too much? Mike |
11th Feb 2014, 6:39 pm | #1056 |
Octode
Join Date: Mar 2006
Location: Watford, Hertfordshire, UK.
Posts: 1,270
|
Re: The "Sussex" Homebrew Valve Tester.
Hi Mike, the two resistors are from the original RAT valve tester circuit.
I believe Mike Rowe did not want to re-invent the wheel and used parts of the RAT schematic. The 1M and 10k provided a tap that was used for the the mu measurement in the original article, as shown on attachment. Les P.S. I guess the 10k could be dispensed with, but the circuit board carries both.
__________________
Whether the Top Cap is Grid or Anode - touching it will give you a buzz either way! |
11th Feb 2014, 6:59 pm | #1057 | |
Nonode
Join Date: Aug 2010
Location: Resolfen, Wales; and Bristol, England
Posts: 2,588
|
Re: The "Sussex" Homebrew Valve Tester.
Quote:
Just my 2 cents.... Richard |
|
12th Feb 2014, 12:27 am | #1058 |
Rest in Peace
Join Date: Mar 2011
Location: Basildon, Essex, UK.
Posts: 4,100
|
Re: The "Sussex" Homebrew Valve Tester.
Hi
I checked the screen voltage and that does the same it builds up, a bit faster than I said before, about 2 seconds to full voltage, the anode is the same. I have taken some more traces with an EL33 valve under test. The first image is a trace of the sawtooth waveform again, this time a little less, it builds up over a couple of seconds from nothing to its peak value as shown. Connections to scope were from ground and HT side of fuse using x10 probe. The second picture is taken with connections to ground and the fuse, this trace starts from nothing as soon as the HT is switched off by the relay, and builds up to this strange wave form over 2 seconds. Mike |
12th Feb 2014, 9:37 am | #1059 | |
Rest in Peace
Join Date: Mar 2011
Location: Basildon, Essex, UK.
Posts: 4,100
|
Re: The "Sussex" Homebrew Valve Tester.
Quote:
Thanks Mike |
|
12th Feb 2014, 10:11 pm | #1060 |
Rest in Peace
Join Date: Mar 2011
Location: Basildon, Essex, UK.
Posts: 4,100
|
Re: The "Sussex" Homebrew Valve Tester.
I tried the above and it made no difference.
Please can someone help me and explain where this sawtooth waveform is coming from, if it is normal, and if not normal, how to eliminate it. Could you also please confirm if the type of problems I have been experiencing with EL34 valves is caused by a lack of ferrite beads. If so what beads should I use, where would I get them, and how are they fitted. Thanks Mike |