DIY something like iMPAMP??

[aletheian]

wow, you guys are a trip! 8^)

ROTF!!!

 

[userguide]

HEY ! Welcome to the DiyAudio forum Zachary.
You are now a member of our esteemed community.

Six months it is a long reaction time to a post you know...

So could you share with us more on your ImpAmp design ?
I don't obviously want to affect your business, just eager for our knowledge and all great ideas to be shared.

 

[zachary vex]

well, it's got two independent linear power supplies for the high voltage on each channel. it has the max run on the rails on the OP275 chips, the bias for the tubes is a blue LED. it's got way-over-engineered transformers (capable of 5W), the washers on top of the board are for magnetic isolation to avoid coupling between the toroids and the output transformers, i use a TI rail-splitter to create the ground for the entire circuit, and there's no signal paths longer than 1.5 inches in the entire amp, completely eliminating any possibility of internal buzz or ground-loops.

other than that, there's not much to say. oh, other than that i use a magic trick to create the rail voltage for the preamp chips. 8^)

 

[pinkmouse]

Welcome Zachary, it's always nice to see people here that actually do this lark for a living, especially with such a neat product as yours.

Be aware that these guys will pester you silly for info, (in a good natured way, of course), so be firm with them!

So to start the ball rolling, I am very interested in your idea of using a rail splitter for grounding. I'm not asking you to reveal your implimentation, but I would be very interested in any references you have to such a topology.

 

[aletheian]

well, it's got two independent linear power supplies for the high voltage on each channel. it has the max run on the rails on the OP275 chips, the bias for the tubes is a blue LED. it's got way-over-engineered transformers (capable of 5W), the washers on top of the board are for magnetic isolation to avoid coupling between the toroids and the output transformers, i use a TI rail-splitter to create the ground for the entire circuit, and there's no signal paths longer than 1.5 inches in the entire amp, completely eliminating any possibility of internal buzz or ground-loops.

other than that, there's not much to say. oh, other than that i use a magic trick to create the rail voltage for the preamp chips. 8^)

i remember watching the demo videos on the website a while ago and being impressed by the inginuity it took to squeeze a stereo PP triode amp like that into a smaller hammond 1590-sized box, REGARDLESS of the wattage. I got bitten by the sub-mini tube bug a few months before the product introduction and actually bought an array of various minis, but then lost interest before anything got past the computer simulation stage. I was working on PP pentode stages anyway.

 

[A. J. Gandulfo]

6BX7 ANDREI AMP

Dear Andrei, congratulations for your clean design.

Can you, pleas tell me wich plate to plate load selected and wich is the plate voltage and idle current selected for the 6BX7 output stage.

Thanks for your replay.

Regards

 

[A. J. Gandulfo]

6BX7 Andrei Amp

Dear Andrei

Yesterday I make some simulations with Broskie software package and your schematic.

The 6BX7 maximun disipation is 12W, both sections combined and with these limit for me, the best sceneario is:

Plate Voltage= 270V
Idle Current= 20mA / Triode
Cathode Resistor= 560 Ohm
Total disipation at idle= 9.5W
Heater Disipation= 9.5W
Total Disipation= 19W
Plate to Plate Load= 8 KOhm
LOad Impedance= 6 Ohm
Output Impedance= 4 Ohm
Power Output @ 2% Third Harmonic Distortion= 5W
Input Signal= 24VPeak / Grid
Efficency= 25%
Maximun Power Output= 9W @ 4.5% Third Harmonic @ 32VPeak / Grid
Maximun Peak Plate Voltage= 480V

The simulated transformer is:
Primary Impedance= 8 KOhm
Speaker Load= 6 Ohm
Efficency= 97%
Primary resistance= 300 Ohm
Secondary resistance= 0.4 Ohm

The first 6SN7 Stage:

Gain= 14 with 100 KOhm Load

May be the Tamura 10K transformer is not the best selection because the frequency range is 300Hz @ -2dB
ElectraPrint in their website, have the same application with the 6BX7, but using their U$S 51 choque, I hope with more inductance, the recomended minimun load for the 6SN7 may be around 25 / 30KOhm, and using mathematics for 20Hz the primary inductance @ 20V RMS is 200Hy.
ElectraPrint use another tube for driver, the 6J6 or in another schematic a Pentode.

Please let me know your comments

 

[andrei]

Those numbers are pretty close to my simulations. Transformers I'm using are commodity parts - cheap and easily available. 8k plate-to-plate load might be a bit easier for 6BX7, but unavailable in Amveco's power toroid lineup. Same goes for a small Tamura. Although I'm using a damped resonance to extend the low-frequency responce, it still is not too impressive (cuts off at 100Hz or so). There probably also will be some saturation-induced distortion for low frequency signals at full volume.

You can get better results by using better transformers (input transformers are a good idea, too), but then the amp will be bigger, heavier, and more expensive. Your call .

 

[forsakenrider]

I know this is an old thread but i have some 6111's and I would like to build something like this.

did anyone ever build this amp? or does anyone have any ideas for a low watt amp I can build with these tubes. (not a headphone amp, ive already built a few).
I want to build something I can put in my kitchen with some small 8ohm book shelve speakers

 

[erix]

Bringing this thread back from the dead...

This is not exactly like the iMP but it is close. I've got it running on a breadboard as I type...
Differences are many:
- Uses 12AU7A instead of 6021 (seems to not mind 12AT7, 12AX7, 12BH7)
- Different power supply (more caps than iMP)
- Different circuit around inverting opamp stage (compared to schem in iMP video)
- Many more that I know and don't know...

Seems to be running fine with a bench speaker - have to bring it upstairs for a real speaker load.

Only thing it doesn't do have iMP does (at least on all the videos I've seen) is throbbing LED. The LED in mine stays constantly bright meaning the tube bias isn't exactly right. A resistor between the tube cathodes and the LED will probably be required.

Sorry for the zip file - I don't have photo-editing software on this computer!

 

[revintage]

This is an interesting thread. But why use OPamps as drivers?

I would use a 6H30 selfsplit with currentmirror and half a 6DJ8 as driver. I have a tiny 15VA 115+115:6//6 Talema to try out. Will do a sim in LTSpice and come back.

Could be used together with this little creature of mine and built in a similar 120*120*32mm box:

 

[revintage]

The 15VA Talema is 8,8k:8ohm. In PP this gives us 4W of class-A power with cathode bias, Ua 200V, B+215V and Iq 30mA/tube. To build it two Hammond-boxes will be needed, one amp and one PSU.

 

[userguide]

The 15VA Talema is 8,8k:8ohm. In PP this gives us 4W of class-A power with cathode bias, Ua 200V, B+215V and Iq 30mA/tube. To build it two Hammond-boxes will be needed, one amp and one PSU.

Would you care to share your schematic?

 

[revintage]

Will share when they are ready in LTSpice. So far I have only been using my:

www.revintage.se/PPABAMP.xls


One schematic with the above mentioned combination and one with 4 6C45 and LL1544A 1:2 input transformers.

Planned the amp yesterday and the complete amplifier an be built in two boxes of same type as the preamp with the stereo poweramp in one and the PSU in the other.

 

[erix]

Some measurements from the breadboard:
AC supply: 13VAC
Filament: 10.3VAC
B+: 200.8V
Plate Current: 9.9mA (with 470r resistor in series with LED)
V+: 16.04VDC
V-: -15.98VDC

Sounded pretty decent until I broke a pin off the 12AU7A...

Correction to my post above: It does NOT like 12AX7, but sounds really REALLY nice with 12BH7A.

" But why use OPamps as drivers?" heh, why not?

Actually it's an academic exercise: more CAN it be built rather than SHOULD it be built....

 

[revintage]

Seems like the whole project is an academic exercise. With max 500mW out it is otherwise a total waste of time and money !

If you bias it to 25mA(max for the 125A) you will get well over 1Watt out.

 

[erix]

20mA = max plate current of 12AU7A!

 

[revintage]

Hi userguide,
This is the schematic I promised. Will not try the 6C45 version.

http://www.revintage.se/PPA.PNG

Simmed it will deliver 3W below 1% THD and 4W clipped/compressed at 5% THD.

The network at the input (together with C4) is to compensate for a pair of single driver Tangband OBs I am building and can be removed. For linear response C4, C1 and C2 should be adjusted to 1u and 2200u.

Sensitivity 1,4Vrms for 3W/8ohm at 1k w/ filter , 0,6Vrms w/o filter

 

[revintage]

20mA = max plate current of 12AU7A!

So your indicated plate current was for each half? In that case no problem, just make the 470ohm slightly smaller so you get 2*12,5mA.

 

[erix]

well....
That was the current for both halves I guess.
I'll measure the current of just one - seems I might have more current available than I thought!