Hi Rich,
Following your last post I went back and followed your suggestions. As usual I took photos of the oscilloscope traces for the various set-ups. (See attached pdf file).
Some details for the photo's are given below.
Fig 1: No zobel on primary of 125GSE output trafo and no negative feedback circuit. Used ~ 0.13V pk to pk 1kHz square wave. This gave an output signal of ~ 4.5V but with strong peak on leading edge of the trace.
Fig 2: No zobel but this time 4k7 resistor in parallel with 220pF cap was used. Due to the drop in output voltage with the neg. feedback in use the input voltage of the square wave was increased to ~0.27V pk to pk for similar output. This time trace showed strong peak (slightly stronger than fig 1) with some ringing.
Fig 3: No zobel but feedback loop now 4k7 in parallel with 470pF Cap (I couldn't find a 330pF). Everything else is the same as for fig 2. There is a slight reduction in the magnitude of the initial peak.
Fig 4: No zobel but feedback loop now 4k7 in parallel with 2200pF . Everything else is the same as for figs 2,3.
Now the large cap has virtually eliminated the strong peak at the start of the trace.
So, without Zobel across primary of output trafo, it appears that a large cap in combination with feedback resistor is required to produce a satisfactory square wave but I'm not sure this is a step forward see Miles Prower quote from an earlier post.
So what do you recommend? Just for completeness I have a photo with zobel in use + feedback loop also in. See next post!
Cheers
Colin.
Following your last post I went back and followed your suggestions. As usual I took photos of the oscilloscope traces for the various set-ups. (See attached pdf file).
Some details for the photo's are given below.
Fig 1: No zobel on primary of 125GSE output trafo and no negative feedback circuit. Used ~ 0.13V pk to pk 1kHz square wave. This gave an output signal of ~ 4.5V but with strong peak on leading edge of the trace.
Fig 2: No zobel but this time 4k7 resistor in parallel with 220pF cap was used. Due to the drop in output voltage with the neg. feedback in use the input voltage of the square wave was increased to ~0.27V pk to pk for similar output. This time trace showed strong peak (slightly stronger than fig 1) with some ringing.
Fig 3: No zobel but feedback loop now 4k7 in parallel with 470pF Cap (I couldn't find a 330pF). Everything else is the same as for fig 2. There is a slight reduction in the magnitude of the initial peak.
Fig 4: No zobel but feedback loop now 4k7 in parallel with 2200pF . Everything else is the same as for figs 2,3.
Now the large cap has virtually eliminated the strong peak at the start of the trace.
So, without Zobel across primary of output trafo, it appears that a large cap in combination with feedback resistor is required to produce a satisfactory square wave but I'm not sure this is a step forward see Miles Prower quote from an earlier post.
So what do you recommend? Just for completeness I have a photo with zobel in use + feedback loop also in. See next post!
Cheers
Colin.
Hi Rich,
Just a continuation of the last post:
Attached is a photo of
Zobel of 4k7 in series with 1.8nF cap. Feedback circuit R= 4K7, C= 220pF. Horizontal = 0.2ms/div. Vertical scale for 1kHz square wave = 0.1V/Div, Vertical scale for Output signal = 1V/Div.
So it appears to be a compromise now:
1. Use zobel + 4k7/220pf.
2. Forget zobel, use 4k7/220pf-470pf and accept slight 'ringing'?
What in your opinion is the best way forward now?
Thanks for all the help over the last few days - it's much appreciated.
Cheers
Colin.
Just a continuation of the last post:
Attached is a photo of
Zobel of 4k7 in series with 1.8nF cap. Feedback circuit R= 4K7, C= 220pF. Horizontal = 0.2ms/div. Vertical scale for 1kHz square wave = 0.1V/Div, Vertical scale for Output signal = 1V/Div.
So it appears to be a compromise now:
1. Use zobel + 4k7/220pf.
2. Forget zobel, use 4k7/220pf-470pf and accept slight 'ringing'?
What in your opinion is the best way forward now?
Thanks for all the help over the last few days - it's much appreciated.
Cheers
Colin.
Attachments
ColinB;
2200pF is an overkill as Miles comments will ruin HF response as will excess Zobel across primary.
Compromise is use 470pF max on global nfb loop and light zobel.
Thin rise and fall flanks on waveforms indicate high speed. Ideal for good HF response.
The slight ripple on these vertical rise and fall flanks could be caused by
1; camera reflection ??
2; check your scope probe grounding. A lanky ground will reflect square wave rise and fall. Connect close to 2ndary windings.
If scope is earthed through same grounding as power amp via AC mains earh then a conflict can arise through grounds. Remedy, use isolation tranny on amp. Moving the probe cable around is a clue and if the waveform quality varies then only one ground is required. (the scope)
3; If above isn't the problem then in some amps increasing primary current can oblivate the ripple above problem. This isn't advised unless there is a sound reason to do so.
Comment 2 is serious consideration when getting into the big power game when high dV/dt really starts to tax ones experience.
RF experience is a huge advantage.
I'm tempted to leave things at that and shove music through the amp.
richj
2200pF is an overkill as Miles comments will ruin HF response as will excess Zobel across primary.
Compromise is use 470pF max on global nfb loop and light zobel.
Thin rise and fall flanks on waveforms indicate high speed. Ideal for good HF response.
The slight ripple on these vertical rise and fall flanks could be caused by
1; camera reflection ??
2; check your scope probe grounding. A lanky ground will reflect square wave rise and fall. Connect close to 2ndary windings.
If scope is earthed through same grounding as power amp via AC mains earh then a conflict can arise through grounds. Remedy, use isolation tranny on amp. Moving the probe cable around is a clue and if the waveform quality varies then only one ground is required. (the scope)
3; If above isn't the problem then in some amps increasing primary current can oblivate the ripple above problem. This isn't advised unless there is a sound reason to do so.
Comment 2 is serious consideration when getting into the big power game when high dV/dt really starts to tax ones experience.
RF experience is a huge advantage.
I'm tempted to leave things at that and shove music through the amp.
richj
Hi Rich,
Thanks for your reply.
I think the main culprit could be due to camera motion whilst taking pictures without flash. They do look narrower on the scope display prior to photography.
I will check the values or R & C in the zobel tonight as I can't remember whether I fully optimised the values or not.
After this I'll test with music with Zobel in and 4k7/470pf on feedback loop in a temporary set-up using clip leads etc so I can tweak if necessary.
From the scope traces it looks like the input sensitivity of the power amp is going to be pretty high so I guess I'll use my MP3 player (with lower output signal rather than CD player with output at 'line' level.)
Anyway, I just want to thank your for help, patience & support over the last few days. I sure I wouldn't have made it work without your help
Cheers
Colin.
Thanks for your reply.
I think the main culprit could be due to camera motion whilst taking pictures without flash. They do look narrower on the scope display prior to photography.
I will check the values or R & C in the zobel tonight as I can't remember whether I fully optimised the values or not.
After this I'll test with music with Zobel in and 4k7/470pf on feedback loop in a temporary set-up using clip leads etc so I can tweak if necessary.
From the scope traces it looks like the input sensitivity of the power amp is going to be pretty high so I guess I'll use my MP3 player (with lower output signal rather than CD player with output at 'line' level.)
Anyway, I just want to thank your for help, patience & support over the last few days. I sure I wouldn't have made it work without your help
Cheers
Colin.
Hi Rich,
Thanks for asking how the amp sounds...
Firstly, a quick summary of where I'm at with the amp. After some minor tweaking I started with
1. Zobel (3k3 +1000pF)+ 4k7/470pF feedback loop. Then removed Zobel. To be honest with my 'worn-out' ears I couldn't really hear any significant differences with Zobel in or out.
So started to listen to music with only 4k7/470pF feedback in place using initially my MP3 player (which allows me to switch between lots of songs pretty quickly). One thing I noticed immediately was that there was lots of treble energy and generally not much bass. (I was using my MP3 player > preamp (5687) > ECL82 Power amp.) I assumed this was probaly down to the relative poor source component.
Next I switched to CD but of course I hit the usual problem of too much gain. I couldn't get the volume knob past 9 o'clock - just too loud. I've been in this situation before with other valve amps but this time I decided I make up an interconnect with a pair of resistors (6k8 & 3k3) in the phono plugs to give ~10dB attenuation. This allowed me to reach ~11 o'clock with the volume knob - slightly better. In addition, I reduced the feedback resistor to 3k3 & retained the 470pF cap which also helped reduce the excessive gain.
After further listening, however my overriding impression is still of a dynamic but bright almost 'thin' sound from both sources. I must admit that the cheap Hammond 125GSE output trafo's are likely to be bass shy anyway so this won't help but at the moment I'm undecided whether this amp will stay or in the long term be taken apart to build something else - further listening is required.
There is some buzz & hiss audible when close to the speakers but I have built amps that have worse background noise than this one. (The buzz probably comes from my pre-amp which I've never been able to cure. The slight hiss I suspect is from this amp (ECL82).
In the last year or I've built 4 amps + the one above (without knowing much about valve theory behind them). These were
1. DC Darling (8532 + 1626 tubes)
2. Bevois Valley (Morgan Jones) EL84 PP
3. Gary Kaufman 6EM7 DC coupled amp
4. EL34 SE amp.
Out of the five I would say that the Bevois Valley has probably the best sound but perhaps only marginally so.
Regards
Colin.
Thanks for asking how the amp sounds...
Firstly, a quick summary of where I'm at with the amp. After some minor tweaking I started with
1. Zobel (3k3 +1000pF)+ 4k7/470pF feedback loop. Then removed Zobel. To be honest with my 'worn-out' ears I couldn't really hear any significant differences with Zobel in or out.
So started to listen to music with only 4k7/470pF feedback in place using initially my MP3 player (which allows me to switch between lots of songs pretty quickly). One thing I noticed immediately was that there was lots of treble energy and generally not much bass. (I was using my MP3 player > preamp (5687) > ECL82 Power amp.) I assumed this was probaly down to the relative poor source component.
Next I switched to CD but of course I hit the usual problem of too much gain. I couldn't get the volume knob past 9 o'clock - just too loud. I've been in this situation before with other valve amps but this time I decided I make up an interconnect with a pair of resistors (6k8 & 3k3) in the phono plugs to give ~10dB attenuation. This allowed me to reach ~11 o'clock with the volume knob - slightly better. In addition, I reduced the feedback resistor to 3k3 & retained the 470pF cap which also helped reduce the excessive gain.
After further listening, however my overriding impression is still of a dynamic but bright almost 'thin' sound from both sources. I must admit that the cheap Hammond 125GSE output trafo's are likely to be bass shy anyway so this won't help but at the moment I'm undecided whether this amp will stay or in the long term be taken apart to build something else - further listening is required.
There is some buzz & hiss audible when close to the speakers but I have built amps that have worse background noise than this one. (The buzz probably comes from my pre-amp which I've never been able to cure. The slight hiss I suspect is from this amp (ECL82).
In the last year or I've built 4 amps + the one above (without knowing much about valve theory behind them). These were
1. DC Darling (8532 + 1626 tubes)
2. Bevois Valley (Morgan Jones) EL84 PP
3. Gary Kaufman 6EM7 DC coupled amp
4. EL34 SE amp.
Out of the five I would say that the Bevois Valley has probably the best sound but perhaps only marginally so.
Regards
Colin.
It doesn't surprise me that your amp sounds a bit thin. My personal experience of the ECL82 is that it is a very bright sounding tube. My speculation is that they were voiced that way to help compensate for the poor treble response of cheap radio output transformers, of course I could be wrong.
It took me three attempts to get this tube to sound good in a circuit, and now I have it in a little PP headphone amp. Sounds really good. I have also seen it used in a very expensive amp using an interstage transformer.
Unfortunately if you don't like the result its really a matter of stripping down and trying something else until it does sound good.
Shoog
It took me three attempts to get this tube to sound good in a circuit, and now I have it in a little PP headphone amp. Sounds really good. I have also seen it used in a very expensive amp using an interstage transformer.
Unfortunately if you don't like the result its really a matter of stripping down and trying something else until it does sound good.
Shoog
Hi Colin,
From your first post I see that you have built and used Gary Kaufman's 6EM7 DC coupled triode amp with some Fostex 167E speakers. Being a newbie, and with only one solo build so far, the Boozhoundlabs excellent 6V6 SET amp, I have already begun collecting components for Gary's 6EM7. This hobby is really addictive!
I have constructed some speakers using the Fostex 167E drivers, see www.geocities.com/rbrines1/Pages/Proposals.html and am very pleased with the results. I use my speakers in a large room but they fire across the width so I don't need a lot of volume and I mainly listen to fairly light female and male vocal music.
I would appreciate any comments or tips you may have regarding your pairing.
Thanks Graham
From your first post I see that you have built and used Gary Kaufman's 6EM7 DC coupled triode amp with some Fostex 167E speakers. Being a newbie, and with only one solo build so far, the Boozhoundlabs excellent 6V6 SET amp, I have already begun collecting components for Gary's 6EM7. This hobby is really addictive!
I have constructed some speakers using the Fostex 167E drivers, see www.geocities.com/rbrines1/Pages/Proposals.html and am very pleased with the results. I use my speakers in a large room but they fire across the width so I don't need a lot of volume and I mainly listen to fairly light female and male vocal music.
I would appreciate any comments or tips you may have regarding your pairing.
Thanks Graham
Hi Graham,
I've been pondering for some time what to say about the Gary Kaufmann 6EM7 amp.
The pro's are it is an easy amp to build and when completed I had no hum or background noise and I only used cheap components from Maplins e.g. paralled coffin resistors to get target 1.5K and 1.3K and Jamicon 47uF caps.
For the power supply I used Hammond 370HX power trafo followed by 47uF Evox-Rifa cap (this was part of a power supply from an aborted EL34SET) followed by JJ 47uF + Hammond 157L choke for each channel. Output trafos were Hammond 125GSE.
It was all put together in the chassis that had previously been the EL34SET amp.
However, for me the amp didn't really produce the goods with my Fe167E speakers - it wasn't 'loud' enough because the amp is low powered (~1W) and the FE167E speakers are not particularly efficient (94dB). I reckon a more efficient speaker is needed here (98dB or more). I decided to take it apart after only 1 week's listening. I've had a similar experience with my DC Darling that I built prior to the 6EM7 - good for background listening but just not loud enough for 'serious' listening.
However,
you may well be ok with mainly vocal type music - it's certainly worth building as it is not too challenging and is pretty cheap so I would so go ahead with it.
By the way here is link to an 6EM7 discussion on a UK DIY site (you may have already seen this).
http://www.audio-talk.co.uk/phpBB2/viewtopic.php?t=185
I used my FE167 in the standard Fostex recommended cabinets which from reading posts here (@fullrange forum) is probably not the best option. I had considered trying a MLTL type but my experience of building the original cabinets put me off this for now.
Anyway, good luck with your build.
Cheers
Colin.
I've been pondering for some time what to say about the Gary Kaufmann 6EM7 amp.
The pro's are it is an easy amp to build and when completed I had no hum or background noise and I only used cheap components from Maplins e.g. paralled coffin resistors to get target 1.5K and 1.3K and Jamicon 47uF caps.
For the power supply I used Hammond 370HX power trafo followed by 47uF Evox-Rifa cap (this was part of a power supply from an aborted EL34SET) followed by JJ 47uF + Hammond 157L choke for each channel. Output trafos were Hammond 125GSE.
It was all put together in the chassis that had previously been the EL34SET amp.
However, for me the amp didn't really produce the goods with my Fe167E speakers - it wasn't 'loud' enough because the amp is low powered (~1W) and the FE167E speakers are not particularly efficient (94dB). I reckon a more efficient speaker is needed here (98dB or more). I decided to take it apart after only 1 week's listening. I've had a similar experience with my DC Darling that I built prior to the 6EM7 - good for background listening but just not loud enough for 'serious' listening.
However,
you may well be ok with mainly vocal type music - it's certainly worth building as it is not too challenging and is pretty cheap so I would so go ahead with it.
By the way here is link to an 6EM7 discussion on a UK DIY site (you may have already seen this).
http://www.audio-talk.co.uk/phpBB2/viewtopic.php?t=185
I used my FE167 in the standard Fostex recommended cabinets which from reading posts here (@fullrange forum) is probably not the best option. I had considered trying a MLTL type but my experience of building the original cabinets put me off this for now.
Anyway, good luck with your build.
Cheers
Colin.
Hi Colin,
I'm so glad you pondered over your reply as you have given me a lot to consider.
As you recommend, I think I will continue to construct the 6EM7 amp as I have puchased the valves (via E-bay USA).
Your PSU details are also going to be handy as I was about to take a look at the Duncanamps PSU design programme.
That forum link is a new one to me but I recognise a lot of the guys from the World Designs forum.
Anyhow, if my speakers struggle a little with the low output, it's a great excuse to build a more efficient design!
Many thanks for your assistance.
Graham
I'm so glad you pondered over your reply as you have given me a lot to consider.
As you recommend, I think I will continue to construct the 6EM7 amp as I have puchased the valves (via E-bay USA).
Your PSU details are also going to be handy as I was about to take a look at the Duncanamps PSU design programme.
That forum link is a new one to me but I recognise a lot of the guys from the World Designs forum.
Anyhow, if my speakers struggle a little with the low output, it's a great excuse to build a more efficient design!
Many thanks for your assistance.
Graham
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.