Aleph J illustrated build guide

So I'm safe to carry on with bias procedure? I no longer have 24v across outputs.

I tend to use the bulb tester just for testing the power supply in isolation.

Then I use a lab power supply with current limiting to set the bias and dc offset of the amp, then I back off on the bias and connect it to the amp power supply to make the final small adjustments.

That probably doesn't help you right now. Best of luck.
 
I think ZM secret recipe for motor run cap is very amazing, AJ can show it's benefit. Even with a cheap one can give you extra smile 😀
 

Attachments

  • IMG20160426213710.jpg
    IMG20160426213710.jpg
    463.3 KB · Views: 924
I read through this thread off and on for months but never spotted specific tolerances for matching JFets and Mosfets.

I have a run of 8 Mosfets that tested VGS from 3.3227, 3.3243, 3.3243, 3.3265, 3.3267, 3.3273, 3.3274, 3.3289

Are these good enough to drop in anywhere? I don't know how tight my tolerances should be or if outputs are more important etc...

Also on the Jfets I have a run of SJ74's from 6.48, 6.55, 6.72, 6.74, 6.81, 6.90... Are these good enough? I have other matched pairs but this is the closest I have in a foursome. I was about to solder the upper 4 in when I read that the differential jfet matching was "critical". How critical is critical? Do I put a 6.72 and 6.9 on one channel and a 6.72 and 6.81 on the other or match pairs per channel?

Some feedback on this would be nice. I'm dropping fets in and then all boards are done PSU and Trafo are wired up and about 60% with my chassis so have another month or so... I've invested hundreds of dollars and hundreds on man hours in this so wanna get it right. Figured the doubt warranted asking before I do something stupid.
 
Throw Down ......

mosfets - well , you can throw down last two numbers in all of them , and declare them perfectly matched

go for it (hopefully they're matched at or close to Iq destined in circuit say that 500mA is minimum for that test)

JFets - as memory serves me - you need pairs (for LTP) , nothing bigger (triplets or quads)

no need that they're hysterically matched between channels

and yeah - for mosfets - you need matched pairs per rail ; so - one matched pair up , second matched pair down
 
The mosfets are fantastically close. That's a matched Octet if ever there was one. Use wherever you choose without further thought.

Those Jfet are also quite closely matched - usually within 10% is close enough for current sharing, you've got it.

As for 'balance' between channels, you want the devices as close to each other as possible, not the best average between L and R ... in your devices, 6.48, 6.55 on one board, 6.72, 6.74 on the other. But honestly, you could blindly choose from the pile and have no issues.
 
I am rerunning my test

Throw Down ......

mosfets - well , you can throw down last two numbers in all of them , and declare them perfectly matched

go for it (hopefully they're matched at or close to Iq destined in circuit say that 500mA is minimum for that test)

JFets - as memory serves me - you need pairs (for LTP) , nothing bigger (triplets or quads)

no need that they're hysterically matched between channels

and yeah - for mosfets - you need matched pairs per rail ; so - one matched pair up , second matched pair down

So my critters got into my parts bin and knocked a few mosfets off the foam boards I had them stored in and I need to rerun the test.

I originally used a 100r and a 15v supply which -4v for VGS was around 110ma... So if I retest with the same 15v supply I should use a 22r to get current up to 500ma?
 
So my critters got into my parts bin and knocked a few mosfets off the foam boards I had them stored in and I need to rerun the test.

I originally used a 100r and a 15v supply which -4v for VGS was around 110ma... So if I retest with the same 15v supply I should use a 22r to get current up to 500ma?

22R sounds about right. Just be careful of the dissipation though.
I would use a 20W resistor there.

Cheers,
Dennis
 
Yeah I dont have a 20 watter laying around so I'll need ti hit the electronics shop on Tuesday.

Grimberg, the 8 matched fets were on a foam board with 32 others. The kids knocked the foam board off its perch and 3 matched and 2 unmatched came loose. I didn't mark up the fets so have no way of knowing which is which now.

I could just test the 5 but figured if I'm going to build a jig again I may as well go ahead and test as per the circuit at 500ma.

This time around I'll have a silver sharpie in hand so this doesn't happen again.
 
JT Electronics in Westminster. They don't have a lot of quality stuff. Not gonna find Alps or Bourne or Nichicon or ELNA per se, but they have a fair amount of generic stuff. I know they have 20 watt power resistors as long as I'm not looking for Vishay or Ohmite etc...

I did get 100 TO-247 mica pads that neither Mouser or Digikey have from them. Plus a couple nice heat sinks from their surplus/ clearance aisle.
 
Thinking of new projects...

I already have some projects going on and they seem to take longer to finish the closer to finish I come....
Could it be something about searching for new ideas and projects that come in the way??😉

Anyway, I have been looking into this interesting Aleph J as a follow up to the tube amp that is to be finished and the Aleph P1.7 that is in pipeline and the....and....

I started to think about the PSU for the Aleph J and I know that many here have used the PSU from the shop but now it is so that checked some old boxes and I found some 10 pieces of Panasonic 15000 uF 63 V caps that should be working for the Aleph J. Unfortunately they measure 40 mm diameter so they will not fit the DIYShop board but I then thought that I could wire the power supply point to point.

20160804_115457.jpg

The schematic show the use of 4 caps per rail with two before and after the 0.47 resistors.
I was thinking to use all 10 of the caps that I found so here is the question, do I place three before the 0.47 resistors or do I put three caps after the resistors???:scratch1::scratch1:
What is the advantage of doing one or the other and why?😕

It would be a shame to leave two resistors when I can use the uF:s better in a Aleph J. The rails would then have 75000 uF on each rail and that would be better than the stipulated 60000 uF.
Somewhere I read that "excess is good"..... I go for that!😀



Just now listening to Nils Landgren, Sentimental Journey