Just to show how committed and nice Daniel is, he PMed me seeing if I got my 7293 para. kit up and running yet! Talk about service! 
I haven't yet due to a divorce move but plan on it in the next month or two.
Oh and Daniel, your PM box is full and won't accept any more messages--I can now understand why it's full!
Cheers,
Jeff
I haven't yet due to a divorce move but plan on it in the next month or two.
Oh and Daniel, your PM box is full and won't accept any more messages--I can now understand why it's full!
Cheers,
Jeff
Sorry to hear about the divorce. Audio too loud? Anyway, I had a couple of thoughts. . .
A number of bookshelves could be more convenient than so many boxes. Thanks for the compliment! The TDA7293 Parallel (if modded as indicated), the LM1875 Parallel (not using datasheet sample values), the TDA7297 Bridge (prefab and convenient), all of those have the same pretty tone for LOW listening fatigue.
One could choose them based on room size, large, medium, or small:
Great Room style house: TDA7293 Parallel
Traditional style house: LM1875 Parallel
Apartment: TDA7297
Speaker efficiency could be another way to choose.
A number of bookshelves could be more convenient than so many boxes. Thanks for the compliment! The TDA7293 Parallel (if modded as indicated), the LM1875 Parallel (not using datasheet sample values), the TDA7297 Bridge (prefab and convenient), all of those have the same pretty tone for LOW listening fatigue.
One could choose them based on room size, large, medium, or small:
Great Room style house: TDA7293 Parallel
Traditional style house: LM1875 Parallel
Apartment: TDA7297
Speaker efficiency could be another way to choose.
A little off-topic but still about using these amplifier modules....
I will soon complete a scratch build of the Transcendent (Bruce Rozenblit) 300B SE-OTL amplifier; this uses four 300B tubes per channel in an output transformerless design. Straight from the schematic it uses 50/60Hz AC filament heating but Transcendent use a 25Hz filament supply in their kit version which means hum is made inaudible and, interestingly, is supposed to improve the sound quality. In due course I would like to try a 25Hz filament supply but don't want to spend to much and thought a couple of these modules, fed from a sine wave generator might do the trick. The potential problem is the demand that supplying the 300B filaments with a 25Hz sine wave would place on each amplifier module (assuming one per channel); four 300B filaments would represent a load of a bit under 4ohms and require a constant 30watts - is that going to be too much for them?
I would appreciate your views.
Ta muchly
Ray
I will soon complete a scratch build of the Transcendent (Bruce Rozenblit) 300B SE-OTL amplifier; this uses four 300B tubes per channel in an output transformerless design. Straight from the schematic it uses 50/60Hz AC filament heating but Transcendent use a 25Hz filament supply in their kit version which means hum is made inaudible and, interestingly, is supposed to improve the sound quality. In due course I would like to try a 25Hz filament supply but don't want to spend to much and thought a couple of these modules, fed from a sine wave generator might do the trick. The potential problem is the demand that supplying the 300B filaments with a 25Hz sine wave would place on each amplifier module (assuming one per channel); four 300B filaments would represent a load of a bit under 4ohms and require a constant 30watts - is that going to be too much for them?
I would appreciate your views.
Ta muchly
Ray
For very low bass pitches, TDA7293's can do 45W each, and you've got 2 of them, so 90W, 4R.
Info: Inbuilt limiter "skims" low bass pitches when power exceeds 45W per each chip (so I estimate 90W in our case).
For 25hz, you'll need to oversize the coupling caps, such as positive input 4.7uF (to input the bass pitch) and inverting input 1000uF (to ALLOW amplifying the 25hz bass pitch as cleanly as possible--actually oversizing the inv-input coupler tilts the response to prefer lowest bass rumbles rather than midbass booms)
The decoupling caps need to be somewhat 470uF (although I much prefer 220uF||220uF = 440uF as photographed in the mods on this thread). At some loss of mids and treble clarity (for this particular chip), you can max out the decouplers with 330u||330u, without causing extra heat. Values larger than 470uF solo or 330uF dual for the decouplers may cause a great deal extra heat and loss of performance (the datasheet is wrong). To address the heat issue further you can put 1 (one) 2.2uF 250v cap (either polyester or Mallory SEK electrolytic is fine) from V+ to V- at the amplifier board, for a slight bit more cool running.
I've actually no idea about your application, but this bit can be used for a subwoofer amp with a very deep voice.
Regarding power estimate, if harshly biased, you'll be needing the very similar triple-parallel board, but if class ab, the 2-chip parallel board in this thread will do fine.
Last edited:
This bit may be unclear:
Maximum amplifier board power decoupling cap per each rail is:
470u for single capacitors (per each rail) <-rather easy to install
330u||330u for parallel capacitors (per each rail) <-preferred for sub amp
220||220u for parallel capacitors (per each rail) <-preferred for full band
Notes:
*However, for subwoofer duties (and similar), you can get the power supply board really really close to the amp board and use 12ga solid copper for DC umbilical. Power Supply for subwoofer is at least 13200uF per each rail, and Not a CRC. Multiple smaller caps (such as 3300u) are preferred. Twin KBPC3502 rectifiers recommended (voltage drop pattern is like a straight up wall = doesn't cause sag).
**If using the diode trick photographed earlier, but with a subwoofer amp, use MBR1645 (an excellent low loss part) for series elements at the amplifier board (instead of the 6a05's photographed, MBR1645 substitute is to decrease loss; however, the filtration still works, as does the huge dynamic wallop).
***If using a CRC power board with subwoofer amplifier, again use the MBR1645, but this time, in parallel with the CRC's "R" in order to prevent loss of power at the CRC's resistor (also prevents some filtration, but not all of it).
Question:
Ray, I don't know anything about tubes; however, if a 25hz filament supply sounds great, then wouldn't a 0hz filament supply (an ordinary voltage regulator chip boosted for current by an amp output transistor) sound better?
P.S.
Other use of this board is a 90W buffer, done by assembling as indicated, and then re-working the master chip into slave mode (kills the voltage amp and you get zilch gain). Then one applies global negative feedback from any preamp, driver chip, tube preamp, etc...(replacement voltage amplifier driving the chip's predrive input pin). The small signal triode (like a 5670, 6922, etc...) can do the job, and then one can apply variable current drive adapter from Lenard Audio to imitate an output transformer sound. Biggie size tubes not required and neither are output transformers. This optional mode can be used to make an extremely high resolution hybrid amplifier at minimal cost. I'm curious if Aikido can drive it. If dual mono and aikido, then you've got enough tubes and transformers "on show" and plentiful high resolution audio (and excessive power). That may be worth a try.
The chip's inbuilt voltage amp either wants the gain too high or the voltage too low, thus makes either a high power party blaster with a nice tone or a flea power analytics amp with fantastic imaging, but, like most chip amplifiers, does not do high end imaging while also at full blast. Yes, nearly that, but not quite. It may be fun to experiment with a replacement voltage amplifier, and quite visually attractive if that is tubes. Cheers!
Details/data:
Maximum amplifier board power decoupling cap per each rail is:
470u for single capacitors (per each rail) <-rather easy to install
330u||330u for parallel capacitors (per each rail) <-preferred for sub amp
220||220u for parallel capacitors (per each rail) <-preferred for full band
Notes:
*However, for subwoofer duties (and similar), you can get the power supply board really really close to the amp board and use 12ga solid copper for DC umbilical. Power Supply for subwoofer is at least 13200uF per each rail, and Not a CRC. Multiple smaller caps (such as 3300u) are preferred. Twin KBPC3502 rectifiers recommended (voltage drop pattern is like a straight up wall = doesn't cause sag).
**If using the diode trick photographed earlier, but with a subwoofer amp, use MBR1645 (an excellent low loss part) for series elements at the amplifier board (instead of the 6a05's photographed, MBR1645 substitute is to decrease loss; however, the filtration still works, as does the huge dynamic wallop).
***If using a CRC power board with subwoofer amplifier, again use the MBR1645, but this time, in parallel with the CRC's "R" in order to prevent loss of power at the CRC's resistor (also prevents some filtration, but not all of it).
Question:
Ray, I don't know anything about tubes; however, if a 25hz filament supply sounds great, then wouldn't a 0hz filament supply (an ordinary voltage regulator chip boosted for current by an amp output transistor) sound better?
P.S.
Other use of this board is a 90W buffer, done by assembling as indicated, and then re-working the master chip into slave mode (kills the voltage amp and you get zilch gain). Then one applies global negative feedback from any preamp, driver chip, tube preamp, etc...(replacement voltage amplifier driving the chip's predrive input pin). The small signal triode (like a 5670, 6922, etc...) can do the job, and then one can apply variable current drive adapter from Lenard Audio to imitate an output transformer sound. Biggie size tubes not required and neither are output transformers. This optional mode can be used to make an extremely high resolution hybrid amplifier at minimal cost. I'm curious if Aikido can drive it. If dual mono and aikido, then you've got enough tubes and transformers "on show" and plentiful high resolution audio (and excessive power). That may be worth a try.
The chip's inbuilt voltage amp either wants the gain too high or the voltage too low, thus makes either a high power party blaster with a nice tone or a flea power analytics amp with fantastic imaging, but, like most chip amplifiers, does not do high end imaging while also at full blast. Yes, nearly that, but not quite. It may be fun to experiment with a replacement voltage amplifier, and quite visually attractive if that is tubes. Cheers!
Last edited:
Daniel, thanks for taking the time to reply with such comprehensive information; I think I'll order a pair of the kits and give it a try. At worst I'll lose a few pounds from my pocket.
I need to reread the thread too.
Do you know if the power supply kit offered as an option on this listing up to the job?
TDA7293 x2pcs 170W Mono amplifier board Kit -29 | eBay
or can you point me in the direction of a suitable PSU.
I think the speaker protection module might be good too, and/or something with a soft start, just to make the life of the filaments as easy as possible.
To answer your question; yes, conventionally a DC heater supply would be an option but a key element of the design eliminates that possibility. There is also a whole separate debate around which is better for powering filaments in direct heated triodes, AC or DC, but this isn't the place nor I the person. I just want to try the different options.
Ray
In this case, a scrap of copperless phenolic perfboard, 6 of 6800uF caps (use 3 per rail), a pair of prefab bridge rectifiers (use 1 per secondary) and some lengths of 12ga solid copper cable. I'd also add mild drainers and add some LED's to the rails for safety indication.
Note: The CRC function isn't needed for subwoofer amp. Some power supply examples That page is out of date--those are older models, but they do work. Problem: The output diodes (MBR1645) and their ferrites should actually go at the "amplifier end" of the dc umbilical cable. Those parts are optional but allow maximum performance amplifier board decoupling to work. Even with a subwoofer amp that is useful.
P.S.
My schematic may be hard to read with that 680R+50VR, so here it is: The gain divider is 730R vs 27k. Option: Use a 1k multi-turn cermet trimmer as a variable resistor for the feedback-shunt resistor so you can dial it around for best behavior (about 730R). In that case, the feedback resistor is still at 27k. See post 1.
Last edited:
Slower start? Possibly upsize the mute and standby caps to 47u and or decrease resistor current, may make the amp come on more slowly and more gently. Or it may burst instead. Or it may simply not turn on. I really don't know. But, it seems likely that some option like that could slow down the engagement.
Alternative: A voltage divider at the output of your tone generator, using an LDR, with its led in parallel to a capacitor, the capacitor charged through a resistor, would almost certainly make a slower start. Use the resistor normal for led current, and upsize the cap to slow it down a lot.
Possibly also doable with a jfet instead of the LDR.
Last edited:
A cap Is needed on the standby.
A cap is not needed on the mute.
It is okay to reduce the turn on click by using the mute cap; however, I took the mute cap off of my TDA7293 in order to reduce the number of RC's in the power circuit. There was almost no audible difference either way.
It doesn't take 4 parts to operate the mute (2 resistors, 1 diode, 1 cap)--Actually, all it takes to switch the mute is 1 resistor within approximately 10k~25k range.
A cap is not needed on the mute.
It is okay to reduce the turn on click by using the mute cap; however, I took the mute cap off of my TDA7293 in order to reduce the number of RC's in the power circuit. There was almost no audible difference either way.
It doesn't take 4 parts to operate the mute (2 resistors, 1 diode, 1 cap)--Actually, all it takes to switch the mute is 1 resistor within approximately 10k~25k range.
For the Anteks, did you go with the AN or AS series. Thanks for your vendor suggestions as a result of which I have a kit on the way.
it simply means "quality sound"
perhaps one of the sellers can identify the designer for you
DC filter: See mod at post 58 and yet more photos earlier. That will reduce power noise considerably. This is also good for imaging.
AC filter: To further reduce power noise, snub the transformer secondaries with RC's. This is also good for imaging.
Doing both results in the two of the most power filters (except regs) available for an audio amp, and would surely null the difference between the AN or AS transformer.
P.S.
Also, with both AC and DC filters mentioned above, you won't need a CRC power board, but a rather plain KBPC2504+5x3300uF per rail, would do an excellent job.
if you don't mind a clarification, I gather you're saying that the feedback shunt R is 1k5 and both the feedback and input Rs are 47k. so the BOM for the changed parts, in sum, is:
2- 47k
1- 1k5
1- 220uf
1- 150uf
yes?
That should indeed result in a working amplifier.
Although it is quite different from mine.
However, I have left an interesting schematic at post#1 (the start of this thread), which lists the component values that I'm actually using.
If you have two boards available, it may be interesting to build one amp one way and the other amp the other way, so that you can compare and see which you like.
Basically, the 27k and 730R feedback divider sets high gain (also quite stable) and is suitable to higher voltage; however, the 47k and 1k5 feedback divider sets lower gain and is suitable to lower voltage.
The DC filter photographed earlier, works well for both.
Although it is quite different from mine.
However, I have left an interesting schematic at post#1 (the start of this thread), which lists the component values that I'm actually using.
If you have two boards available, it may be interesting to build one amp one way and the other amp the other way, so that you can compare and see which you like.
Basically, the 27k and 730R feedback divider sets high gain (also quite stable) and is suitable to higher voltage; however, the 47k and 1k5 feedback divider sets lower gain and is suitable to lower voltage.
The DC filter photographed earlier, works well for both.
Last edited:
Due to the well working filters (added to the amp board with mods) shown in the photo attachements at Post#30, there is no need for a CRC type power supply board. Therefore, for the documentation at Post#1, I'm uploading an edited power supply photo that is simpler.
This is 7x2200uF per rail and it has two bridge rectifiers made from 8 diodes
(Or much easier with 5x3300uF per rail and a pair of prefab bridge rectifiers.)
This is 7x2200uF per rail and it has two bridge rectifiers made from 8 diodes
(Or much easier with 5x3300uF per rail and a pair of prefab bridge rectifiers.)
Attachments
just a little heads-up that there is a very minor discrepancy in post #1 between the schematic and the text (they differ as to the input load R value). i doubt that anyone would be led astray by this but i thought it worth mentioning since so much else you've done so carefully and thoroughly.
Last edited:
FWIW--looks like the same designer did this 3xchip version--has the same chinese symbols on the board: TDA7293 x3 BTL 250W Mono Amplifier Kit 26 | eBay
Last edited:
