[JohnBedini1]

Mr T
Thanks for responding privately by E-mail.
Here are the pictures you asked for. This is one of the first LM317 K Amplifiers, we keep it upgraded here with the new caps and transformers. We use this amp everyday playing music of all kinds, it is sort of are reference amp. The only rival amp is the model 10/10 Class A. Was glad to give you a little history. unfortunately I can not up load pictures here so here are the links. I know your friend wants to stay silent through all of this but glad he is a member here. I did not want to do this, but he must understand that I have been involved in a broad energy physics field since the time I was involved with  talking to John Iverson about electromagnetic research of magnetic motors, and no I do not know where his motor has gone. Anyway it should look just like the one you have only with a screen top. As your friend knows What keep everybody ignorant is condemnation before investigation,
Scope Shot as you asked for Square wave at 1KHz http://www.johnbedini.net/john34/scopeshotsq.jpg
The picture of the Amp  http://www.johnbedini.net/john34/lm317k.jpg
Hope this helps.
John B

 

[T-Rexx]

John Bedini,
Thank you so very much for all your help.. As you know I'm in the mountains above you but I did not bring the amp. I would like to come and visit you next week if you do not mind. My friend thanks you too. As you know I can not have a computer in CA because of my job. We will talk. Again thank you from the both of us for the pictures.
Take Care
Mr T

 

[nikongod]

I built a simple amp using the LM317 (or a depletion mosfet) in CCS mode driving a small output transformer. Well, its stereo so theres 2 of everything.... Think of a single stage transformer coupled tube amp. Now replace the tube with an LM317. I wired a resistor in parallel to the primary of the transformer to effectively drop the Zo of the LM317=CCS.... Bandwidth was not good (at all) without that. 
 
It was a fun project, and I have thought extensively about building an "OTL" amp using nothing but LM317, caps, and resistors. Maybe a few LEDs for good measure. Everyone likes LEDs. 

 

[Avro_Arrow]

  Everyone likes LEDs. 

LEDs are priitttyy

 

[JohnBedini1]

nikogod,
Sorry I do not know if your addressing me or someone else here. But if you are, then it is very intriguing what your saying about the LM317 when used as an amplifier. I never considered using a transformer with it. I always built power followers as can be seen in my picture. This was discussed in the late 70's by a number of engineers on how to build a high powered LM317 amplifiers. This is one chip that has many functions if you study it's internals. As used in audio it has a good bandwidth if the measurements are correctly done. I don't have any problem running the chip at  10Hz to 150Khz into any load, however the "noble network" is very important, the loading must be correct at any frequency. THD measurements are always very good .00012 it is usually at the limits of the analyzer. When using Spice unless the models are correct it will not conform to the bench method, and you will find yourself recalculating it over again as you try to get it to work. The unit we have here runs any load we put it on, so I'm happy with it's performance, great on headphones.
John B

 

[nikongod]

Hi John
 
My post was just kind of genericly addressed. Seemed like an appropriate thread for it and all. 
 
Here is a photo of the amp, and the original schematic with the DN2540:
 

 

 
I originally built the amp with the DN2540 but *NOT* the 650ohm resistor not considering the effect the very high output impedance of the FET would have on the bandwidth of the transformer. After a bit of head-scratching as to why the amp started to roll off at 2khz i assumed something must be wrong with my FET (I dont know why) and replaced it with a LM317. The B+ voltage is low enough that it is fine. After the LM317 gave the SAME bandwidth I added the 650ohm resistor and bandwidth went to the raitings of the transformer. 
 
The only issue (I consider it a feature - its a matter of perspective) is that the voltage gain of the amp is less than 1. Since I primarily use this amp for Grados, and I generally listen quietly it does not bother me much at all. 

 

[JohnBedini1]

nikongod,
Let me just say a few things that were discussed between John Iverson and I before he had gone missing. It was at a CES show, I was in his room talking about amplifiers and the LM317. He said something that caught my attention. It was about class AB and how the answer is in that design. He called the Eagle 7A the bitch amplifier but after listing for a while I caught what he was talking about. He suggested that I should and try to couple the LM317K device to the output of that amplifier (7A) I did not think about at the time. Now I can see what he was saying after allot of work with that device.I have devised a way to do that I just can't post anything here or change my Avatar for some strange reason.
It seems you have put allot of work into your amplifier with very expensive transformers, I have not tried that but I'm going to. So your using this for head phones right? I use the amp here to run the Shahinian Diapasons bridged. The low impedance is around 1.6 ohms at the low frequencies. I have not tried the stax phones yet as I'm waiting for box to plug them into. But all the other phones work well with it. I wish I could post some pictures but I can not as of yet. The thing is if your happy with what you have built and it sounds good do not change it. Sometimes measurements mean nothing when it comes to music .25% could sound good. I'm going to try a new type design with this device, have not seen it anywhere as of yet. I use the T03 type devices which is the K device, the heat sinks are 4X10  so I have plenty of heat sink still runs around 125 Degrees but does not go over that. I will post as soon as I can.
John BE

 

[nikongod]

Yes, Im usiong the amp with headphones. I'm not sure if there is a way to get enough gain/voltage out for speakers even with a gain stage. Garry Pimm has gotten lots of power out of a similar idea but with much higher voltages.

The transformers in my amp were pretty inexpensive. The total parts cost was about $100 and you could go a bit below that with surplus power supplies (or just a bench supply).

I think you need to post more before the forum let's you upload images - to prevent/slow spam. Just post "test" until you get past 10 or 20 (not sure which) posts and it let's you load pics. I can't wait to see the schematics.

 

[wakibaki]

Reading John Broskie's TubeCAD blog from 2003, I came across this...
 

 
...since I see I can get these for ~$0.30 from Thailand, I think I'll work up a PCB on the CNC mill. Only thing is ~1.2A per channel. A gain of 5 would probably be plenty.
 
w