Well done insanity.  Bodes well for the listening test to come.  Hope you like the musical quality

 

And thank you Kevin and Birgir I like the KGST very much.

It is a cracking amplifier.  Next time around I'll test one with the current production pentodes. 

Could you elaborate what you mean by cracking?

Could you elaborate what you mean by cracking?

 

It's slang. Usually means something very good. Or to make the word good even more so. 

 

http://www.urbandictionary.com/define.php?term=cracking

Today I tried to listen to some music for the first time. One channel is perfectly fine. The other one only transmits low frequencies with distortion and no high frequencies. Without a signal it is dead silent. I changed the tubes, no effect. Where would you start to look for the problem? I guess it must be some device in the 1st or 2nd stage maybe?

Edited September 19, 201411 yr by insanity

Can you adjust the balance and offset on the offending channel? 

Yes everything adjusts fine. 

How is a bias voltage? Too high or low...?

Bias voltage is from a same circuit but if diaphragms have a different tolerance.

Bias is set to 580V, 

Do you have another pair of headphones to try?  If you rewire the channels so that the old right is the new left (and vice versa), does the problem swap to the other channel?  I'm just thinking about ruling out the possibility of a bad pair of headphones.  Do the headphones work fine in another amp?

Edited September 19, 201411 yr by chiguy

The headphones work fine on my kgsshv. I also switched the source channels to rule out a problem with the source. 

Tomorrow I will double check every connection and do a comparative measurement of both channels hoping to find a difference.

Edited September 19, 201411 yr by insanity

Maybe the diode testing method will help you locate where the problem is. See Kerry's post at The ultimate DIY? A Stax SRM-T2! This have helped me a couple of times.

Ok guys something weird is going on. I replaced the input stage of the "faulty" channel  - no effect. Then I switched the O+ and O- wires running to the stax jack from one board to the other and vice versa. I expected the problem to switch sides. But the problem stays on the same side! WTF? 

As I said the headphones work perfectly with the KGSSHV. If I interpret this correctly, this leaves me with either a bias problem or a faulty cable to the stax jack. 

 

Just to be sure how to correctly set the bias voltage. Measure between GND and Bias Testpoint and turn the 1M pot until = 580V. Measurement after the 4.7M resistor gives me around 400V. 

If I interpret this correctly, this leaves me with either a bias problem or a faulty cable to the stax jack. 

 

 

Or maybe a connection at the Stax Jack itself?

Measure the offset/balance with headphones connected and then without, there could be some flexing of a cold join somewhere?

Ok guys, I am an idiot. I messed up the wiring of the stax jack. I switched the bias and one of the output lines...

I don't really know how it was possible to get good sound on one side like this.

Everything is fixed now, works as it should. I feel so ashamed.

 

Sound seems pretty good so far, but I cannot really comment much on difference to the KGSSHV yet. I will try to do this, once it is cased up and connected to my regular source. This will take some time, since I have not yet decided on a case to get.

use the little stax jack circuit boards. makes it easy.

I have those... But I didn't want to put the pcb on the jack yet. With the PCB mounted it seems impossible to insert screws. 

How do you guys do it. Screw the jack into the case and then mount the pcb?

I use a circular PCB.

I mount the sockets to the chassis and there they will stay. 

the square ones work fine with my stax jacks.

 

the round ones work regardless.

 

best to mount Justin's jacks to the front panel first, or drill

all the way thru the front panel and use screws and nuts

and then you can take it off.

Thanks for sharing your experiences. Since I have the PCBs from Birgir, I have to wait to mount them until the chassis is ready.

 

When I was looking at the KGST and most recent KGSSHV boards, I asked myself the following question:

Since these boards incorporate the 2 100ohm input resistors which are directely in the signal path before amplification, would there be any profit in choosing higher quality resistors than the usual rn55 oder xicons? Of course the signal passes many other resistors on the board as well. But this was just a thought. 

I was tossing around the idea of mounting Stax jacks using threaded rods and nuts using blind holes on the back of the panel.  My other thought was cap screws also with blind holes.  

When I was looking at the KGST and most recent KGSSHV boards, I asked myself the following question:

Since these boards incorporate the 2 100ohm input resistors which are directely in the signal path before amplification, would there be any profit in choosing higher quality resistors than the usual rn55 oder xicons? Of course the signal passes many other resistors on the board as well. But this was just a thought. 

 

The signal passes through a lot of the other resistors too...    Stick with RN60D and it will be fine. 

 

For the record, I added those resistors to help prevent attenuator issues.  They can be bypassed but I'd recommend keeping them as they won't do any harm and can prevent some minor headaches. 

I know what they are for. I still have to add them in the KGSSHV though. 

 

A question to running KGST owners: Are you amps completely silent like the KGSSHV or is there a minimal amout of background noise? After some more listing to mine today, which is still uncased and a provisional setup, I noticed that its not dead silent after all. Its not a grounding issue - its not a classical 50hz hum. It's more of a higher pitched sound. 

 

Is it normal that there are some unpleasant sounds when turning off the amp with the headphones plugged in?

 

Here is a pic of my test setup: 

It should be dead silent.  Did you clean the flux off the boards?  That could be creating a short somewhere.