I've been completing the construction of my hybrid power amp with an E88CC tube as a buffer. I expect to have it finished in a couple of days, do some final measurements and then I plan on doing some listening tests and compare it with the small power amp Aleph3 I built some time ago. I could present the results of those tests here if there's enough interest. First, a few pics:
Regards
Regards
Attachments
pots
I've noticed all the IGCs have pots. Is the point of the pot to eliminate the need for a preamp?
I don't really have a need for a pot on the amp as I wish to continue using my preamp and ulitmately active crossovers. Is it difficult (for an amateur) to remove? Is driving it with a preamp a problem? I have a Foreplay but I'm reading Audio Reality and I'm thinking about buiding the GG pre.
I've been eyeing this project for a while and love the idea of combining tubes with ICs.
Thanks.
I've noticed all the IGCs have pots. Is the point of the pot to eliminate the need for a preamp?
I don't really have a need for a pot on the amp as I wish to continue using my preamp and ulitmately active crossovers. Is it difficult (for an amateur) to remove? Is driving it with a preamp a problem? I have a Foreplay but I'm reading Audio Reality and I'm thinking about buiding the GG pre.
I've been eyeing this project for a while and love the idea of combining tubes with ICs.
Thanks.
Re: pots
You don't actually need the pot IF the IGC is buffered - indeed you can then use it as a Power Amplifier, so wire it straight into an RCA socket. Instead of the pot put a 47K or 100K resistor.
The gain is quite high, in this DIY Tube Gainclone the gain is 33dB and that means a preamp is superfluous. Do as moamps has done (see his wonderful pics) and you can see his input selector switch, I think it looks like 3, wired before the pot, and... you have a INTEGRATED amplifier.
The original IGC will not work properly without the pot, but this one will, so it's up to you.
Joe
ultrachrome said:
You don't actually need the pot IF the IGC is buffered - indeed you can then use it as a Power Amplifier, so wire it straight into an RCA socket. Instead of the pot put a 47K or 100K resistor.
The gain is quite high, in this DIY Tube Gainclone the gain is 33dB and that means a preamp is superfluous. Do as moamps has done (see his wonderful pics) and you can see his input selector switch, I think it looks like 3, wired before the pot, and... you have a INTEGRATED amplifier.
The original IGC will not work properly without the pot, but this one will, so it's up to you.
Joe
Re: Re: Re: pots
Good question.
The original IGC had a flaw, one that its originator (hi T) must have realised. When the pot was set at maximum (and hence would equate to your above suggestion), the gain dropped to 17dB. These chips are likely to go unstable at gains below 20dB. In practice, with 2V RMS sources, like most CD players, the pot in the original IGC would unlikely ever to be advanced to its max position. In the mid position the gain was about 27dB (but total voltage gain was less because the pot was in the input circuit) and all was OK. As you can see, gain changed with pot variations.
BUT if the pot was replaced with a fixed 100K resistor, that would be most unwise. Besides, even if the chip was stable at 17dB, the IGC sounds better, IMHO, when the gain is set at slightly more than 30dB.
Joe
PS: The newer version of the IGC - revised by its originator (you-know-who), has gain set so it never goes below 20dB. So he knew about it and fixed it.
pedroskova said:
Good question.
The original IGC had a flaw, one that its originator (hi T) must have realised. When the pot was set at maximum (and hence would equate to your above suggestion), the gain dropped to 17dB. These chips are likely to go unstable at gains below 20dB. In practice, with 2V RMS sources, like most CD players, the pot in the original IGC would unlikely ever to be advanced to its max position. In the mid position the gain was about 27dB (but total voltage gain was less because the pot was in the input circuit) and all was OK. As you can see, gain changed with pot variations.
BUT if the pot was replaced with a fixed 100K resistor, that would be most unwise. Besides, even if the chip was stable at 17dB, the IGC sounds better, IMHO, when the gain is set at slightly more than 30dB.
Joe
PS: The newer version of the IGC - revised by its originator (you-know-who), has gain set so it never goes below 20dB. So he knew about it and fixed it.
Re: Re: Re: Re: pots
This is a quote from a third page of this thread:
>>Let’s look at the way feedback varies with pot changes. The gain is set by values (220K+10K)/10K = 23, or 27.2dB, – but this potential max gain (hence lowest feedback) is reduced by the pot changes. At pot max position the 100K pot is not seen because of ‘Zero Z’ source and no change in gain/feedback. Same applies when the pot is set to minimum (grounded). But what if pot is in the mid position, as this is where it is more to be when in general operation?
The basic maths shows 100K pot will become 25K (because it is ‘Zero Z’ at ground too so it’s 50K in parallel with 50K = 25K) in series with 10K, hence gain (220+10+25)/(10+25) = 7.28 or 17.2dB.
So feedback changes and in the mid-position we have 10dB MORE FEEDBACK.<<
So I understand you meant that in a mid position of the pot, the gain drops by 10db, as this makes more sense. In the light of that, using 100k input to ground resistor shouldn't be a problem.
Joe Rasmussen said:
This is a quote from a third page of this thread:
>>Let’s look at the way feedback varies with pot changes. The gain is set by values (220K+10K)/10K = 23, or 27.2dB, – but this potential max gain (hence lowest feedback) is reduced by the pot changes. At pot max position the 100K pot is not seen because of ‘Zero Z’ source and no change in gain/feedback. Same applies when the pot is set to minimum (grounded). But what if pot is in the mid position, as this is where it is more to be when in general operation?
The basic maths shows 100K pot will become 25K (because it is ‘Zero Z’ at ground too so it’s 50K in parallel with 50K = 25K) in series with 10K, hence gain (220+10+25)/(10+25) = 7.28 or 17.2dB.
So feedback changes and in the mid-position we have 10dB MORE FEEDBACK.<<
So I understand you meant that in a mid position of the pot, the gain drops by 10db, as this makes more sense. In the light of that, using 100k input to ground resistor shouldn't be a problem.
Re: Re: Re: Re: Re: pots
Except that it cannot be guaranteed to be stable at 17dB gain, but IF it is in your sample it may not be in the next?
Thorsten aka KYW has correctly stated so himself. After I posted this info he took note of it, don't know if that prompted him? Don't want to be (or need to be) a big head and claim it was me. But, for two reasons I would not run it at less than 20dB gain, the second: I reckon it sounds better at a bit above 30dB (more gain = less feedback). Indeed the Gaincard is set at slightly above 30dB and NS uses 27dB in its own sample circuit that most of us have seen?
But each to his own. Since stablility is related to a number of things, it is behaviour at HF that is the determining factor. Could National Semis be alert to sample or batch variations, installation variations or just simply too conservative when stating 20dB gain limit. Who knows?
But I think we should adjust this parameter by ear, that's my bottom line. The standard opamp thinking is that feedback is purely a mechanism to set gain provided that it is equal to or greater than specified by the manufacturer. BUT we are NOT MECHANICS, we are audiophiles and we don't take things like that at face value, especially not when it involves what we know is a sensitive (in our circles) topic of feedback.
Regards
Joe
Peter Daniel said:
Except that it cannot be guaranteed to be stable at 17dB gain, but IF it is in your sample it may not be in the next?
Thorsten aka KYW has correctly stated so himself. After I posted this info he took note of it, don't know if that prompted him? Don't want to be (or need to be) a big head and claim it was me. But, for two reasons I would not run it at less than 20dB gain, the second: I reckon it sounds better at a bit above 30dB (more gain = less feedback). Indeed the Gaincard is set at slightly above 30dB and NS uses 27dB in its own sample circuit that most of us have seen?
But each to his own. Since stablility is related to a number of things, it is behaviour at HF that is the determining factor. Could National Semis be alert to sample or batch variations, installation variations or just simply too conservative when stating 20dB gain limit. Who knows?
But I think we should adjust this parameter by ear, that's my bottom line. The standard opamp thinking is that feedback is purely a mechanism to set gain provided that it is equal to or greater than specified by the manufacturer. BUT we are NOT MECHANICS, we are audiophiles and we don't take things like that at face value, especially not when it involves what we know is a sensitive (in our circles) topic of feedback.
Regards
Joe
Re: Re: Re: Re: Re: Re: pots
The previous quote from page 3 was yours. All I wanted to point out is that the decrease in gain is in mid position of the pot and not at maximum, as stated recently. So 100K resistor shouldn't be a problem at input to ground if somebody wants to really use it.
That apart, I noticed you are using 1meg feedback resistor. Is there any subjective defference between let's say 220K and 1 meg? I recently installed 300k resistor at feedback and didn't noticed any particular differences, but I like more gain. I'm also using log pot and it work just fine, so I don't really know if linear is any better?
Joe Rasmussen said:
The previous quote from page 3 was yours. All I wanted to point out is that the decrease in gain is in mid position of the pot and not at maximum, as stated recently. So 100K resistor shouldn't be a problem at input to ground if somebody wants to really use it.
That apart, I noticed you are using 1meg feedback resistor. Is there any subjective defference between let's say 220K and 1 meg? I recently installed 300k resistor at feedback and didn't noticed any particular differences, but I like more gain. I'm also using log pot and it work just fine, so I don't really know if linear is any better?
Hi,
In theory a linear pot + law faking R offers a number of advantages over a log pot:
- you can use the better sounding cermet pots.
- using the linear taper will theoritically do away with any tracking differences between channels that often plague log pots.
- add to that the much lower price and wider available range of linear pots.
Cheers,
In theory a linear pot + law faking R offers a number of advantages over a log pot:
- you can use the better sounding cermet pots.
- using the linear taper will theoritically do away with any tracking differences between channels that often plague log pots.
- add to that the much lower price and wider available range of linear pots.
Cheers,
Re: pots
Hi Peter
OOPS, MEA CULPA!!
I posted the above this morning (local time) and took off to work, then I thought about it again in the car... and the penny dropped... Peter is RIGHT!!!
You are right, IF the input is connected to a Lo Z source like CD player, then the 100K to ground no longer has any affect on the feedback/gain. The gain will be around 27dB. Just don't turn in ON without an active Lo Z source, like, make sure your source is turned ON and connected.
Here is my suggestion:
This will be completely stable even IF there is no input, as with the 50K to ground gain = 23dB. With Lo-Z source 33dB.
You can also try tweaking the 1n3 or 1n2 cap to taylor bandwidth limiting, by ear.
Likewise to pedroskova,
Sorry I got it wrong. My Dad's advice was, whenever you make a mistake, the clever thing is to admit it!
But without the pot and 100K to ground, connect FIRST before turning power ON. Better still, may I suggest you use above... up to you. It's completely stable... and you might find it sounds better?
Joe
Peter Daniel said:
Hi Peter
OOPS, MEA CULPA!!
I posted the above this morning (local time) and took off to work, then I thought about it again in the car... and the penny dropped... Peter is RIGHT!!!
You are right, IF the input is connected to a Lo Z source like CD player, then the 100K to ground no longer has any affect on the feedback/gain. The gain will be around 27dB. Just don't turn in ON without an active Lo Z source, like, make sure your source is turned ON and connected.
Here is my suggestion:
An externally hosted image should be here but it was not working when we last tested it.
This will be completely stable even IF there is no input, as with the 50K to ground gain = 23dB. With Lo-Z source 33dB.
You can also try tweaking the 1n3 or 1n2 cap to taylor bandwidth limiting, by ear.
Likewise to pedroskova,
Sorry I got it wrong. My Dad's advice was, whenever you make a mistake, the clever thing is to admit it!
But without the pot and 100K to ground, connect FIRST before turning power ON. Better still, may I suggest you use above... up to you. It's completely stable... and you might find it sounds better?
Joe
fdegrove said:Hi,
In theory a linear pot + law faking R offers a number of advantages over a log pot:
- you can use the better sounding cermet pots.
- using the linear taper will theoritically do away with any tracking differences between channels that often plague log pots.
- add to that the much lower price and wider available range of linear pots.
Cheers,
Hi fdegrove
I use a pot in the JLTi Tube Hybrid Amplifier that is made in Taiwan. It's hard to find out exactly what type of pot it is. Initially suspected it was cermet. What I've been able to find out is that it is supposed to be on a polycarbonate base with silver deposit and a carbon coating on top. It is Log - with a tube buffer (or any buffer) Lin would be impracticable unless using "pot + law faking R" as you say.
Like you say, Lin pots generally have one advantage, channel balance is usually much easier to get right or be right. But these Logs I'm using have as good balance as Noble Log pots (which are pretty). If some don't I reject them.
They are also a little cheaper than Nobles.
BUT they sound better than Nobles, which I in turn prefer to any of the Alps including Blues and big Blacks. Mind you the latter is very nice too. But these new (they have only been made for a few years) polycarbonate/Silver/Carbon/Logs are available in 50K & 10K.
I buy them in decent quantities, if any of you guys are on PayPal, then we could work out a way for some of you guys to try it out?
Joe
TGC measuring results
Hi all,
I finished my TGC amp a few days ago, and here are the results.
First, I wish to thank everyone for their support and interest. (Especially to Joe for this thread )
On construction:
When I was thinking about building an amp, I wanted to use as many readily available resources as possible and as less money as possible so I piled up parts from discarded equipment and stuff one would expect to find in the back drawers of, say, an audio equipment repair shop. The amp was built on 5 PCBs, designed rather sloppily using an ancient Protel version, then drawn by hand with a permanent marker and etched using HCL and H2O2 (my Protel design is of such a sub-standard quality that it's completely unintelligible and of no use for anybody else unless I sit down and re-design the whole thing).
The box consists of AL plates 2mm thick and 2 heatsinks on each side. The transformer is covered by an Fe lampshade-)) with an opening at its top, which I closed with a brushed round AL plate. That's all.
When I'm working on some project, I somehow always seem to find out useful new information only after I've already fixed the last screw, which is not only frustrating but also means that I would certainly do things differently with the benefit of hindsight. In this particular case, I wish I had made 2 separate mono-blocks and a bit more complex buffer circuitry.
The only serious problem I encountered (and had not anticipated) was a huge amplitude instability (switch on noise) occurring when the amp was being switched on and disappearing once supply voltage stabilised. This was caused mainly by a slow voltage increase on the ECC88 cathode, where the GC input cap (2u2) was connected, partly due to the slow charging of the filter caps, partly due to the slow heating of the tube. It was necessary to insert a "delay-on" circuit that would remove the noise. There were two options (IMHO) - the first one was to insert a relay at the output, which would connect the speakers to the amp output once the amplitude instability disappears; the second one was the shortening of the amp input until the instability stops. I used the first option because the second one needed two relays and it was more difficult to implement.
Hi all,
I finished my TGC amp a few days ago, and here are the results.
First, I wish to thank everyone for their support and interest. (Especially to Joe for this thread )
On construction:
When I was thinking about building an amp, I wanted to use as many readily available resources as possible and as less money as possible so I piled up parts from discarded equipment and stuff one would expect to find in the back drawers of, say, an audio equipment repair shop. The amp was built on 5 PCBs, designed rather sloppily using an ancient Protel version, then drawn by hand with a permanent marker and etched using HCL and H2O2 (my Protel design is of such a sub-standard quality that it's completely unintelligible and of no use for anybody else unless I sit down and re-design the whole thing).
The box consists of AL plates 2mm thick and 2 heatsinks on each side. The transformer is covered by an Fe lampshade
-)) with an opening at its top, which I closed with a brushed round AL plate. That's all.When I'm working on some project, I somehow always seem to find out useful new information only after I've already fixed the last screw, which is not only frustrating but also means that I would certainly do things differently with the benefit of hindsight. In this particular case, I wish I had made 2 separate mono-blocks and a bit more complex buffer circuitry.
The only serious problem I encountered (and had not anticipated) was a huge amplitude instability (switch on noise) occurring when the amp was being switched on and disappearing once supply voltage stabilised. This was caused mainly by a slow voltage increase on the ECC88 cathode, where the GC input cap (2u2) was connected, partly due to the slow charging of the filter caps, partly due to the slow heating of the tube. It was necessary to insert a "delay-on" circuit that would remove the noise. There were two options (IMHO) - the first one was to insert a relay at the output, which would connect the speakers to the amp output once the amplitude instability disappears; the second one was the shortening of the amp input until the instability stops. I used the first option because the second one needed two relays and it was more difficult to implement.
Attachments
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.