2nd Sep 2020, 10:01 pm | #61 |
Octode
Join Date: Mar 2020
Location: Blackburn with Darwen, Lancashire, UK.
Posts: 1,573
|
Re: Hope to start an audio amp with CV415/CV4046/TT15
Well I had a few problems with the amp tonight, very distorted output, volts not correct on the pins, G2 bias on the output valve down to 20 Volts! After replacing a few caps I thought had gone leaky, that had'nt, I put a replacement TT15 Valve in to find it all working again one grid has shorted to cathode!
Anyway a bit late on the tests with the scope comparing input and output. It was set initially at 1KHz sine wave input 360 millivolts P-P giving 8 Volts P-P output into 4 ohms. I then swept from 20Hz up to 1KHz and then in 1KHz steps to 20KHz, not really sure what to expect? The only real phase shift was down at the low frequencies, going up to 20KHz I could detect little change from 1Khz all the way up. Then I did what was probably a daft thing back at 1KHz and stuck a square wave into the input. Anyway pictures here 20 Hz, 1KHz, 10 KHz, 20 KHz, 1KHz square. 20Hz 1KHz 10KHz 20KHz 1KHz square Would appreciate what anyone thinks of the results, what would you have done differently, and if I should do something about the LF shift and how? Cheers Adrian p.s. Just measured the duff valve now cold and have 2Meg between one G1 and G2 ! Last edited by AdrianH; 2nd Sep 2020 at 10:12 pm. |
2nd Sep 2020, 10:17 pm | #62 |
Guest
Posts: n/a
|
Re: Hope to start an audio amp with CV415/CV4046/TT15
Bit of a droop at 20Hz, that is OK, square wave says it all, the slope is the LF droop and no ringing at HF. I would be pleased with those results.
|
2nd Sep 2020, 10:31 pm | #63 |
Octode
Join Date: Mar 2020
Location: Blackburn with Darwen, Lancashire, UK.
Posts: 1,573
|
Re: Hope to start an audio amp with CV415/CV4046/TT15
OK thanks for the comment, guess I put it down to using the Miller/Radford circuit and having a good transformer.
Adrian |
3rd Sep 2020, 6:22 am | #64 |
Dekatron
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 7,670
|
Re: Hope to start an audio amp with CV415/CV4046/TT15
Yep, that's a pretty decent square wave, what's it look like at 100hz & 10khz?
Andy.
__________________
Curiosity hasn't killed this cat...so far. |
3rd Sep 2020, 6:37 am | #65 |
Moderator
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,902
|
Re: Hope to start an audio amp with CV415/CV4046/TT15
You might want to drop a meter in, as a temporary setup to measure the G2 current.
If it gets too high, the dissipation on the g2 structure can be enough to melt things. As the anode voltage dips with signal, g2 can start to look more positive and essentially take over as the target for electron flow. This isn't just the source of the famous tetrode kink, it is also a peak in G2 dissipation. It might just be a bottle that failed because it had a flaw, but it might also be trying to tell you something, so it's best to check. Certainly, with a short inside it, the old valve has no remaining value, so it could be cracked open and autopsied. David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done |
3rd Sep 2020, 8:12 am | #66 |
Octode
Join Date: Mar 2020
Location: Blackburn with Darwen, Lancashire, UK.
Posts: 1,573
|
Re: Hope to start an audio amp with CV415/CV4046/TT15
Hello David, I tapped the glass of the valve with the multi-meter on it and the G1 to G2 resistance cleared, so I am wondering if some cathode coating had flaked and become lodged in the the grids, I did have the amp upside down to connect wires to the input. The valve should have been OK as it came out of an aircraft transceiver, so should have been OK for all orientations?
Can someone explain why square-wave is used, I can see how a transition from vertical to horizontal could represent a high frequency transition, but can not see the point of DC levels in an a.c. coupled amplifier? Adrian |
3rd Sep 2020, 8:29 am | #67 |
Moderator
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,902
|
Re: Hope to start an audio amp with CV415/CV4046/TT15
Cathode flake maybe, but as you said, intended for aircraft use and all orientations. Other possibility is a failed spot weld, but again, this is supposed to be a robust device. So the nastiest possibility is a melted bit of G2. If a flake, you ought to be able to shake it and see it.
I'd do an autopsy. I'd want to know. David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done |
3rd Sep 2020, 9:06 am | #68 |
Octode
Join Date: Mar 2020
Location: Blackburn with Darwen, Lancashire, UK.
Posts: 1,573
|
Re: Hope to start an audio amp with CV415/CV4046/TT15
I put the valve back in to the amp to test and it is working! But not left in there, I would not trust it.
I have just done a few tests as mentioned by Andy 100Hz sine wave first, then 100Hz square. The traces were going off scope so I reduced scope gain and retook the picture. Last one is the 10KHz Square a bit of ringing top and bottom but I notice the test set is not generating perfect square at 10 KHz. 100Hz sine 100Hz square 1 100Hz square 2 10KHz square Sorry for the slightly shaky, blurred pictures I am hand holding the camera. Input signal is always at the top. Adrian |
3rd Sep 2020, 10:30 am | #69 |
Moderator
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,902
|
Re: Hope to start an audio amp with CV415/CV4046/TT15
That 10kHz square is quite good. Also, when you get to these frequencies you can get things like scope probes ringing, so don't automatically assume it's the amplifier without some sort of double check. Amazing how many people use probes and don't check/trim them. Similarly a lump of wire or coax to the scope may need terminating.
The 100Hz squarewaves simply show differentiation due to the AC coupling in the amplifier... both the blocking capacitors and the primary inductance of the output transformer. David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done |
3rd Sep 2020, 11:08 am | #70 |
Octode
Join Date: Mar 2020
Location: Blackburn with Darwen, Lancashire, UK.
Posts: 1,573
|
Re: Hope to start an audio amp with CV415/CV4046/TT15
Thanks David, think I will crack on with the PCB.
Just done a slight change to the footprint for the TT15, need to check the pin spacing for the 7 and 9 pin footprints and then place the order if I can remember how to generate the Gerber and drill files. Adrian |
3rd Sep 2020, 12:42 pm | #71 |
Moderator
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,902
|
Re: Hope to start an audio amp with CV415/CV4046/TT15
Just hold it for a short while. If there really is a problem with the circuit and the suitability of the TT15, you may just be rushing into making a TT15 destroying board. I would definitely want to investigate the failure before going further.
David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done |
3rd Sep 2020, 1:03 pm | #72 |
Octode
Join Date: Mar 2020
Location: Blackburn with Darwen, Lancashire, UK.
Posts: 1,573
|
Re: Hope to start an audio amp with CV415/CV4046/TT15
Too late David. I did it about 20 mins ago!
I did do a few calcs on the screen current, with my lowish 288 Volts and a 27K resistor in the supply line down to a couple of zeners the max G2 current is 11mA if the pin goes short to ground. Under normal operation the current through the resistor is 4mA, the zeners I put there to ensure the volts would not go higher than 175 Volts. The resistor could probably do with being dropped to 22K to give a bit more regulation, but at present the 27K was the only one I had available. And I await information on a better mains transformer higher voltage and more heater amps Getting the boards made to 2oz copper (double thickness) has bumped up the costs quite a bit. The total for the boards is now £23.34, but I got a new business voucher dropping the cost for the boards to £18.87. I have gone for DHL Express priority shipping at £17.97,(there are cheaper) so the total spend on this at present is £36.84 or £7.37 a board. Not a great loss if I completely mess things up. One saving grace is that I have not printed component values on the boards, so changing the odd value will not look out of place. Lets see how things pan out. I need to get some 1.5mm steel sheet to make the chassis. Adrian |
3rd Sep 2020, 1:59 pm | #73 |
Moderator
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,902
|
Re: Hope to start an audio amp with CV415/CV4046/TT15
Ah, the resistor feed ought to save the day. If the screen did try to go high emission, then the resistor would most likely cop it first.
I'm used to high power RF amps where one slight slip usually triggers a spectacularly destructive cascade... though most recently I've been involved with 300W at 1GHz power amps that can be turned on or off in a few tens of nanoseconds... transistors, too! But with these things you learn to be cautious and not go past something that might be trying to give a warning. It might well be some mechanical failure in the internal structure or some wandering trash.. David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done |
3rd Sep 2020, 2:20 pm | #74 |
Octode
Join Date: Mar 2020
Location: Blackburn with Darwen, Lancashire, UK.
Posts: 1,573
|
Re: Hope to start an audio amp with CV415/CV4046/TT15
I think I have to accept it being some trash on the grids, I can not after all this time consider, that the valve may have been treated harshly it was after all from around 1952 (date on a crystal), in the old aircraft set, it has probably seen some service, been thrown around via a few couriers and probably only recently had a bit of respect!
I will keep to one side with a suitable label and only get put in a socket when I do initial tests or have a situation where damaging the valve further would not matter, it did provide some time producing music through the speaker. I seem to have another 3 to go at. If you are dealing with pulsed rf amps you may be able to describe pulsed radar and how they get such high power outputs quoted, perhaps for my radar thread? I have had an email saying my board is now passed to production from approval, so so far so good. Adrian Adrian |
4th Sep 2020, 7:45 pm | #75 |
Dekatron
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 7,670
|
Re: Hope to start an audio amp with CV415/CV4046/TT15
It was bit hard to decipher your 100hz oscillograph, but think it shows core saturation and maybe some RC coupling issues. As mentioned and it looks like to me there's some quite heavy LF attenuation..
"Can someone explain why square-wave is used" I'll try as far as I understand it. Square wave testing shows you frequency response, can show up any stability issues, transient response etc. See - http://www.kennethkuhn.com/students/...ve_testing.pdf & - https://sound-au.com/articles/squarewave.htm & http://www.angelfire.com/ab3/mjramp/sw.html & finally - http://www.r-type.org/articles/art-125.htm I use square waves to select the NFB capacitor, EG use a 5k pot to dial in about the right amount of NFB by taking THD readings and listening on a speaker, then 1khz square wave in, capacitance decade box across the NFB resistor, select C so that the sq wave has the best "squareness", 1khz sq wave in, the same. Flick back and forth till I get the best 1kz sq wave and best 10khz sq wave; there will be a compromise, a point where both as good as you can get. Lastly as David says, SW the coupling to DC, adjust the probe, take reading. Re G2 max diss, I try a 100r Rg2 first then check the voltage drop across it at 0dB/full whack, then adjust as needed taking into account stability. Hope that helped, Andy.
__________________
Curiosity hasn't killed this cat...so far. |
4th Sep 2020, 9:14 pm | #76 | ||||
Octode
Join Date: Mar 2020
Location: Blackburn with Darwen, Lancashire, UK.
Posts: 1,573
|
Re: Hope to start an audio amp with CV415/CV4046/TT15
Quote:
The TT15 is a bit more awkward, the spec sheet says " The maximum permissible DC grid to cathode resistor for fixed bias applications is 100K " Now I know that I am using cathode resistor bias ,so how that changes things I do not know? Quote:
Quote:
Quote:
Each cathode is drawing 24mA with 21.6 Volts across the 900 Ohm cathode resistors so running close to the spec sheet settings. On the G2 current, with the 27K feed resistor, what I can not tell using zeners is how many mA is going through them to ground and how much is going the valve and out via the cathodes resistors, but the spec sheets suggests a 2.6 mA upwards depending on signal, I will have to scope the screen grid to see if it is dipping when being driven. I did ask the transformer manufacturer what the transformer is capable taking current wise and was told 80mA per side quiescent, 120mA peak and being really for the Mullard 5-10 amp with the EL84's I hope I am not going into saturation as I am running less current then two EL84's but comparable volts. Is it volts or current that matters?, I tend to think current as flux is dependent on ampere turns? I find it all interesting, not entirely sure what I am doing at times, but experimentation is fun. I appreciate all the comments. Not sure if I will ever get real Hi-fi specs from it, or if my hearing/brain would notice differences between it all, but it is fun trying to improve things without going daft about it. I will re-run the square wave tests without the input cap and may double up on the 0.1's to the TT15. Would appreciate knowing about the grid resistors in cathode bias for the valve, but suspect it is a suck it and see, go too high a value and risk instability perhaps. Adrian |
||||
4th Sep 2020, 10:52 pm | #77 |
Octode
Join Date: Mar 2020
Location: Blackburn with Darwen, Lancashire, UK.
Posts: 1,573
|
Re: Hope to start an audio amp with CV415/CV4046/TT15
I hope these are clearer pictures, I stuck the camera on a tripod.
100Hz square, note the input is not square, the test set does not give a pure square wave at these frequencies, not really designed for audio testing, the output on the bottom is set for 2 Volts per div. 1KHz Square 10KHz Square, this is also the max the test set will give with a square wave output, sine will go to 20KHz Distortion check just under 5 watts rms. Not really sure that the input cap change and 0.2uF to TT15 valve has made any difference, distortion is perhaps a point of a percent better then a previous test, but I doubt I could ever tell the difference. My dead bug build quality is showing when I clip the probe on some areas such as grid of the TT15 it starts to show instability. I measured the g2 volts and that is stable and steady at 175 Volts, no signs of dipping when running the square wave tests. as to performance, It may be the best it gets. Time to tidy away and do a bit of reading of the links above. Adrian |
5th Sep 2020, 8:33 am | #78 |
Dekatron
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 7,670
|
Re: Hope to start an audio amp with CV415/CV4046/TT15
" I am using cathode resistor bias ,so how that changes things I do not know?" Fixed bias Rg1 is usually about 10 X less, so you can use a larger grid leak with cathode bias.
100k and 0.1u gives you a roll off at 15hz (15.9hz), that's about right but roll off is cumulative, so 1st stage 15hz 3dB 2nd stage another 3dB, so 6dB in total, it's usual I think to roll off at something like 15hz on the 1st stage but make the coupling to the OP stage at lower roll off by using a bigger coupling cap. It's obviously more complex than that as a previous stages ra is in parallel with the following Rg1. I use a sinewave to sq wave adapter for testing at the moment, it uses a 74HC14 I think, details on ESP. For valve amp testing a super fast rise time isn't needed. As you say interesting stuff, keep up the good work & looking forward to seeing your PCB's, Andy. Andy.
__________________
Curiosity hasn't killed this cat...so far. |
5th Sep 2020, 9:07 am | #79 | ||
Octode
Join Date: Mar 2020
Location: Blackburn with Darwen, Lancashire, UK.
Posts: 1,573
|
Re: Hope to start an audio amp with CV415/CV4046/TT15
Quote:
Quote:
The first link you posted was good, I could understand that, will continue with the others later. Cheers Adrian |
||
5th Sep 2020, 9:55 am | #80 | |
Moderator
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,902
|
Re: Hope to start an audio amp with CV415/CV4046/TT15
Quote:
Consequently, a spec for max grid resistor value (meaning grid leak) may be for not letting DC conditions get out of hand. Some RF power valves can get into a thermal runaway condition of their bias currents. So it's safest to accept that bit in the datasheet as good advice, stick to it and increase coupling capacitors as necessary. David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done |
|