What is a reasonable measurement bandwidth to determine performance and identify stability issues on these types of amps?
I can go up to 40kHz with my PC setup. My scope is 100MHz but I will need a good signal generator to go above 40kHz. I really like the pc as it does phase and distortion. With the oscope I am back to the x-y style with limited quantitative.
I can go up to 40kHz with my PC setup. My scope is 100MHz but I will need a good signal generator to go above 40kHz. I really like the pc as it does phase and distortion. With the oscope I am back to the x-y style with limited quantitative.
What type of scope do you have? I'm using a 100Mhz rigol with FFT. I'm going to try the filtered fundamental trick, with my 339a DA to get the resolution up. Then perhaps I can figure out what i'm seeing. I still haven't started building my conditioning board from Pete, for the PC side of things.
Last edited:
I have an old "analog" tek 465B. Have you ordered the parts for the Pete's board? I started building it the end of Dec and still waiting for a couple of chips. Pete helped me get the switches which brings the time frame back to end of March instead of June like originally quoted.
In the meantime I have a manual jig that I use. Most of my measurements are taken with Soundeasy's EasyLab.
I thought it would be important, while we are waiting for iron to arrive, to talk abut what is required to set these amps up to be stable. If someone uses different iron, then the nfb and phase tweeks will be slightly different.
The first step is to get them running well without feedback. i.e. minimum distortion, no motorboating, no HF stuff, and linear phase from 5Hz through ?
In the meantime I have a manual jig that I use. Most of my measurements are taken with Soundeasy's EasyLab.
I thought it would be important, while we are waiting for iron to arrive, to talk abut what is required to set these amps up to be stable. If someone uses different iron, then the nfb and phase tweeks will be slightly different.
The first step is to get them running well without feedback. i.e. minimum distortion, no motorboating, no HF stuff, and linear phase from 5Hz through ?
Well, I understand what you mean, and I agree. The OPT's are at least as important to the final sound quality as the output tubes.
But I bought the KT88's before we decided to re-do our driveway ($5K) and take my aging father on a trip back to the old country (Norway) this summer ($!!K). So yes, the audio priorities are 'conflicted' as we both agree, but life trumps hobbies occasionally. C'est la vie. I can always upgrade later.
[edit] I just discovered the Edcors are $92 ea. I was thinking they were $54 ea., but that was the power transformers. So I may be calling BudP shortly. Thanks for the heads-up. ;-)
..Todd
Last edited:
Here's the BoM and latest schematic. Please let me know more details for the BoM. There are plenty of details missing, and probably some errors.
If you have lots of BoM details to add, just edit the Excel spreadsheet and use a different colour text for your entries so I can keep track of who sent what.
My email address is in the BoM, you can sent updates to me directly if you prefer.
Note: This forum made me ZIP the spreadsheet before uploading. Sorry for the inconvenience. I'll convert to PDF when it's finished. Here's a link to the Google-Doc version (read-only).
..Todd
If you have lots of BoM details to add, just edit the Excel spreadsheet and use a different colour text for your entries so I can keep track of who sent what.
My email address is in the BoM, you can sent updates to me directly if you prefer.
Note: This forum made me ZIP the spreadsheet before uploading. Sorry for the inconvenience. I'll convert to PDF when it's finished. Here's a link to the Google-Doc version (read-only).
..Todd
Last edited:
Silly BoM question...
The 100n caps on each tube's filament contacts to ground are rated on the schematic at 600+ volts. But I cannot imagine why that rating would be required on a 6.3V node. Was that just a typo on my part? There's a big difference in price between qty=16 low voltage 100n caps and 16x 600V caps.
..Todd
The 100n caps on each tube's filament contacts to ground are rated on the schematic at 600+ volts. But I cannot imagine why that rating would be required on a 6.3V node. Was that just a typo on my part? There's a big difference in price between qty=16 low voltage 100n caps and 16x 600V caps.
..Todd
Last edited:
Probably won't need a 600V part here. I ended up using a 400V part cap because I found some that were decent priced. Mouser #539-150104J400EC These are a film type and the price was decent.
I think we could get away with lower voltage.
I think we could get away with lower voltage.
Taj,
NO! - I mean yes!
The heaters sit at some 42V above common - plus +/- 4,5V if you want to be pedantic. The 42V is 'fixed' against spikes by C5 (10µF). So no higher than 70V, which is the lowest working voltage for ceramics I know of. 100V is very common. It depends on your source.
.... and there are not stupid questions, only stupid answers. Chinese say: 'He who ask is a fool for 5 minutes. He who does not ask is a fool for the rest of his life'.
NO! - I mean yes!
The heaters sit at some 42V above common - plus +/- 4,5V if you want to be pedantic. The 42V is 'fixed' against spikes by C5 (10µF). So no higher than 70V, which is the lowest working voltage for ceramics I know of. 100V is very common. It depends on your source.
.... and there are not stupid questions, only stupid answers. Chinese say: 'He who ask is a fool for 5 minutes. He who does not ask is a fool for the rest of his life'.
I forgot we referenced this to a higher voltage. But still, 100V and 250V polypropylene caps are easy to source and cheap enough at this value. I have thousands in bags here (which I suppose is why they were cheap.)
..Todd
Last edited:
Speaking of heaters, On the BoM, I have spec'd the filament transformer at 30VA based on slightly less than 4A draw, thus leaving us 1A+ 'headroom'. Is this correct/reasonable?
..Todd
..Todd
Yes it is reasonable if there isn't other standard VA's avail. Whom ever builds it should measure the voltage when loaded and insert a series resistor to keep voltage within the tube spec. if required.
My kt88 spec sheet says 1.6A each. so 3.2A for the kt88's. 0.6 for the ltp and 0.3/0.6 for input. I come up with 4.1A minimum. Double it if you want stereo.
My kt88 spec sheet says 1.6A each. so 3.2A for the kt88's. 0.6 for the ltp and 0.3/0.6 for input. I come up with 4.1A minimum. Double it if you want stereo.
I'll list it as 30+VA.
SGregory, Your numbers are higher than those listed in Duncan's data. It lists EF86 at 200mA and ECC88 at .365mA. But 5A is still suitable for your values methinks. A 6N1P (6Н1П), on the other hand, draws 600mA if used in the phase splitter, but I don't know if this tube is usable without some changes to the operating points.
..Todd
I used wrong values for LTP and input. I "ASSUMED" that 0.36mA was for each filament in the ecc88.
Trying to keep my build seperate from this thread. Although it does apply to both.
What unit are you specifying. I have a hammond 6.3Vct 30Va at home that I could run some test on this evening.
"Assume" is an acronym don't you know
Trying to keep my build seperate from this thread. Although it does apply to both.
What unit are you specifying. I have a hammond 6.3Vct 30Va at home that I could run some test on this evening.
"Assume" is an acronym don't you know
Last edited:
Taj,
Hang on! I forgot to say that those heater bypass capacitors are actually best as ceramics. That is why I shoved in those low voltages. Ceramics have a poor temp-co, but the C-values are of indirect consequence - there is no 'right' value. The ability to bypass high frequencies is however important; that's where ceramics are best.
I would suggest therefore that you hunt for (disc-type) ceramic capacitors arount the 47nF or higher, some 70V, 100V...... They should be dirt cheap. (Members will also remember that they should be mounted directly on heater pins to short earth - not necessarily the amplifier common, just a lug at the tube socket screw. Any long leads make them of limited value.)
Hang on! I forgot to say that those heater bypass capacitors are actually best as ceramics. That is why I shoved in those low voltages. Ceramics have a poor temp-co, but the C-values are of indirect consequence - there is no 'right' value. The ability to bypass high frequencies is however important; that's where ceramics are best.
I would suggest therefore that you hunt for (disc-type) ceramic capacitors arount the 47nF or higher, some 70V, 100V...... They should be dirt cheap. (Members will also remember that they should be mounted directly on heater pins to short earth - not necessarily the amplifier common, just a lug at the tube socket screw. Any long leads make them of limited value.)
Okay. Thanks again Johan. So X7R or Z5U ceramics are okay for this?
SGregory, I will 'suggest' the appropriate Hammond model for the filament transformer in the notes section. Not sure what model that is, I'm at work at the moment, but it's probably the same one you're using.
Since building a stereo version of this amplifier would introduce a few different complications, and a bunch of different parts, I will avoid even mentioning that scenario on the BoM. It will be based strictly on building a pair of mono-blocks.
..Todd
SGregory, I will 'suggest' the appropriate Hammond model for the filament transformer in the notes section. Not sure what model that is, I'm at work at the moment, but it's probably the same one you're using.
Since building a stereo version of this amplifier would introduce a few different complications, and a bunch of different parts, I will avoid even mentioning that scenario on the BoM. It will be based strictly on building a pair of mono-blocks.
..Todd
Last edited:
Good on the mono. I will test a few conditions and post back the type I have at home. At work also.
The Hammond 166N6 is 6.3V 4A. (25VA), the next size up in the open-frame 166 series is the 166Q6 6.3V@6A.