I will try to find the link to a pretty cool website that has a freeware program that turns your computer into a signal generator (and a few other interesting tools too), It has been a while since I downloaded this program but I will ask another forum where it was and post it here. I need to find it again as I recently upgraded laptops and need to put it on this one too. It basically uses the sound card to produce the signal and you can get a cheap beldon type jack to go from your headphones jack to x2 RCA male ends and plug it right into your amp.
Jeff
Found problem and website
First, the website with the freeware spectrum analyzer/scope/function generator is here: Introduction
I have toyed with it and run the signal generator into a couple of my amps to produce various tones but I have not yet used it in conjunction with a scope. It has a scope too but I understand the bandwidth is limited by the computer processor or something like that....not real sure. Anyway, for those of you who don't have a scope or function generator here is a free one!
I figured out the problem quickly.....I had a bad 5687 tube! Yep, that simple. I had several, what I was told were NOS Tung Sol 5687 tubes and thoughtlessly assumed they were all good. I found there was no voltage on the cathode where you would expect the voltage to be around 4v. Also the B2+ was the same as the B+ indicating no current throught the tube. I swapped out the tube and there I saw the cathode voltage come to 4v and the B2+ drop to 260v. My B+ is running at 290 instead of 280 but it is within acceptable limits. Also there are about 20,000 people without power in my area due to the recent ice storm so I bet line voltages are a bit higher (I should check them).
The amp sound spectacular, better than I expected. The addition of about 500uF on the cathode (bypass) of the 5687 seems to have brought the sensitivity closer to a reasonable 2vrms rather than the 4vrms without it. I still believe it will put out more power with the addition of the CCS that I am waiting on (got a defective batch and had to return them, they are sending me new ones). At the moment it won't even come close to clipping at full volume indicating the input sensitivity is still less than where it needs to be but, other than that it sounds really good.
It has much less power than my KT88 SE I recently finished but it is never-the-less plenty powerful to power my Proac clones to a good listening level.
I am still in the process of building a pair of Sachiko horns, I am sure this thing will rock those.
So, in the end it was a very simple fix, I am almost embarrassed to report it.
Next is a pair of 300B DX JEL monoblocks, the transformers came in the mail today!
Thanks again for everyones input, even though this was a simple fix I always learn from everyones input.
Jeff
First, the website with the freeware spectrum analyzer/scope/function generator is here: Introduction
I have toyed with it and run the signal generator into a couple of my amps to produce various tones but I have not yet used it in conjunction with a scope. It has a scope too but I understand the bandwidth is limited by the computer processor or something like that....not real sure. Anyway, for those of you who don't have a scope or function generator here is a free one!
I figured out the problem quickly.....I had a bad 5687 tube! Yep, that simple. I had several, what I was told were NOS Tung Sol 5687 tubes and thoughtlessly assumed they were all good. I found there was no voltage on the cathode where you would expect the voltage to be around 4v. Also the B2+ was the same as the B+ indicating no current throught the tube. I swapped out the tube and there I saw the cathode voltage come to 4v and the B2+ drop to 260v. My B+ is running at 290 instead of 280 but it is within acceptable limits. Also there are about 20,000 people without power in my area due to the recent ice storm so I bet line voltages are a bit higher (I should check them).
The amp sound spectacular, better than I expected. The addition of about 500uF on the cathode (bypass) of the 5687 seems to have brought the sensitivity closer to a reasonable 2vrms rather than the 4vrms without it. I still believe it will put out more power with the addition of the CCS that I am waiting on (got a defective batch and had to return them, they are sending me new ones). At the moment it won't even come close to clipping at full volume indicating the input sensitivity is still less than where it needs to be but, other than that it sounds really good.
It has much less power than my KT88 SE I recently finished but it is never-the-less plenty powerful to power my Proac clones to a good listening level.
I am still in the process of building a pair of Sachiko horns, I am sure this thing will rock those.
So, in the end it was a very simple fix, I am almost embarrassed to report it.
Next is a pair of 300B DX JEL monoblocks, the transformers came in the mail today!
Thanks again for everyones input, even though this was a simple fix I always learn from everyones input.
Jeff
That was an easy fix. As for the gain; an unbypassed cathode the gain for that stage will be about 10. With the cathode bypassed, and a current source load, it will be about 17. Not quite double, but most of the way there.
Sheldon
What do you mean by RMS ? DC voltage equals its RMS value as it is
I wouldn't use that much capacitance for cathode decoupling, it will cause bias to track slowly which might cause weird distortions. I would adjust it for reasonable cut-off (-3 dB) somewhere in subsonic range, no more than one decade down from the cut-off point of the speakers.
If instead you want fixed bias, use a couple of LEDs (green + red in series should be right in the ballpark for 4V drop).
What do you mean by RMS ? DC voltage equals its RMS value as it is
I am always open to corrections but I think it was even earlier in this thread where we discussed the difference in P-P voltage vs. rms voltage for the input sensitivity of the amp and the difference of the 2. Are we not talking about AC, the input signal is measured in AC, either P-P or vrms, at least that was what was said earlier (maybe a different thread).
Either way, what I am referring to is the input sensitivity of the amp and the ampunt of voltage required to drive it to clipping. Jim Hagerman (creator of this circuit) claimed a 4 vrms input sensitivity without a bypass cap on the driver tube cathode.
Jeff
Did you have a value in mind? I have seen the calculations somewhere, I think Valve Wizzard's website, but I am not sure I would get them correct on the first try. 39uF of the capacitance is done with pp film type and the remainder is with a Nichicon KZ Muse 470uF elctro. I also have 100uF, 220uF Muse caps as well I could replace this 470 with. I even have some 4.7uF Blackgate caps, about 20 of them, I could parallel several of them too.
Jeff
I'm sorry about the RMS part, I read your post and somehow mixed up the references to cathode bias voltage and input voltage (which is evident from my reply). My bad !
Cathode capacitance: with 330 Rk (= this is what the PDF shows) it should be ~100 uF for 5 Hz, which sounds pretty reasonable. 500 uF means -3 dB around 1 Hz instead.
It is my opinion that the cathode bypass capacitor should only be as large as necessary and not any larger. Subsonic signals get amplified just as well as audible sound and can drive output transformer into saturation. This means the output transformer, which should in theory be working perfectly fine at 50 Hz, will for some unapparent reason seem to "lose bass". Suitably small bypass (and coupling) capacitor means lower amplification of inaudible signals throughout the amplifier and therefore fewer potential problems.
Cathode capacitance: with 330 Rk (= this is what the PDF shows) it should be ~100 uF for 5 Hz, which sounds pretty reasonable. 500 uF means -3 dB around 1 Hz instead.
It is my opinion that the cathode bypass capacitor should only be as large as necessary and not any larger. Subsonic signals get amplified just as well as audible sound and can drive output transformer into saturation. This means the output transformer, which should in theory be working perfectly fine at 50 Hz, will for some unapparent reason seem to "lose bass". Suitably small bypass (and coupling) capacitor means lower amplification of inaudible signals throughout the amplifier and therefore fewer potential problems.
Arnulf,
Sounds perfectly reasonable to me. Thanks. While were on the subject, how about the 0.47uF coupling capacitors in this design? Think an 0.22uF would be equally suitable?
I already have 39uF on the bypass with the film pp caps, so I just need another 61uF. I think I have some very nice 47uF Blackgates I could parallel with the films. I wish films were not so large.
Jeff
Sounds perfectly reasonable to me. Thanks. While were on the subject, how about the 0.47uF coupling capacitors in this design? Think an 0.22uF would be equally suitable?
I already have 39uF on the bypass with the film pp caps, so I just need another 61uF. I think I have some very nice 47uF Blackgates I could parallel with the films. I wish films were not so large.
Jeff
Actually your grid resistor is rather low in value so if anything, I would increase the coupling capacitor (up to no more than 1 uF), and I would only do that if I felt the bass is dampened to much. There is an upside to using such a low grid resistance (= lower input susceptibility to outside noise) and compensating for that with slightly larger coupling capacitor isn't a problem. I would definitely not decrease it to 0.22 uF (= 20+ Hz -3 dB) unless it was obvious that OPT was getting into saturation.
Given your current choice of coupling capacitor I would perhaps just stick with those 39 uF if I liked the sound. Why not it with existing (500 uF) capacitance, 39 uF and ~100 uF and see what you like best ? You can then adjust coupling capacitor accordingly, but only change one element at a time.
What kind of performance tradeoff would there be if I wanted to increase the grid resistor to say 1K in order to be able to use what I have on hand already (0.22uF copper foil coupling cap)?
Why does increasing the grid resistor increase the amps susceptibility to outside noise? Are you speaking of the noise from the resistor itself? Or is there some other source of noise we are dealing with here? We are talking about the 330R correct?
Thanks,
Jeff
Why does increasing the grid resistor increase the amps susceptibility to outside noise? Are you speaking of the noise from the resistor itself? Or is there some other source of noise we are dealing with here? We are talking about the 330R correct?
Thanks,
Jeff
I was referring to R14, Rg of V2, which forms a highpass filter in combination with tube's input impedance (huge) and C13 which you want to change.
If you want to keep same (~10 Hz) -3 dB frequency with your 0.22 uF capacitor, double the value of R14 (to 68K).
If you merely replace the C13, the frequency will be 20+ Hz, which is close to the audible range and I definitely wouldn't go that high (3 dB down means 1/2 the signal amplitude). As I said before, one decade down of lowest frequency of interest would be my choice.
If you want to keep same (~10 Hz) -3 dB frequency with your 0.22 uF capacitor, double the value of R14 (to 68K).
If you merely replace the C13, the frequency will be 20+ Hz, which is close to the audible range and I definitely wouldn't go that high (3 dB down means 1/2 the signal amplitude). As I said before, one decade down of lowest frequency of interest would be my choice.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.