Hafler DH-200/220 Mods

Jump to Latest
#1,344
mlloyd1 said:
i'll put on my prognosticator hat for a moment and wager that within 2 days after bob's schematic for the updated hafler appears in this thread, alexmm or someone will have a nice board layout.
🙂

mlloyd1
Click to expand...

Think the secret labratories are still busy, like me.

I try to make Bobs DH-220C to run as DH-120C (original board size in SMD design, where possible) for my two DH-120, together with my preamps DH-110 (BJT design) and my SE-100 (fully JFET design).
I am also working on a DH-220C design with integrated frontend based on 2SK1058 and 2SJ162 ,using the original Hitachi Mosfets and/or the actual Renesas.

A lot work ahead😀
 
#1,345
SK170/2SJ74 compared to LSK489/LSJ689

Hello,

did someone already simulate Bob Cordell suggestion for the DH-220C with matched pairs of 2SK170/2SJ74? Which grade of these is needed?

I know it was Bob's intention to promote these new Linear System's Dual JFETs, but it would be interesting, if the 2SK170/2SJ74 also works and how they perform, compared to the LSK/LSJs.

Best regards
Günni

PS: a first adaption to the PC-25 (for DH-120) from Bobs DH-220C is in layout state
 
#1,346
Your first decision is what current you need the jFET LTP to operate at.
The second decision is at what percentage of Idss do you want your chosen Id to be.
That directly leads to selection of the Idss of the jFETs.
eg.
Let's suppose we want 3mA tail current (for optimum gain) meaning each is operating at 1.5mA
Let's next assume we have evidence that supports operating these jFETs at 80% of Idss for optimum performance.
That leads to selecting 1.9mA Idss devices for the LTP.

Then we review our selection to ensure the devices can never be damaged by excess input signals that might demand massively higher drain currents.
 
Last edited:
#1,347
AndrewT said:
Your first decision is what current you need the jFET LTP to operate at.
The second decision is at what percentage of Idss do you want your chosen Id to be.
That directly leads to selection of the Idss of the jFETs.
eg.
Let's suppose we want 3mA tail current (for optimum gain) meaning each is operating at 1.5mA
Let's next assume we have evidence that supports operating these jFETs at 80% of Idss for optimum performance.
That leads to selecting 1.9mA Idss devices for the LTP.

Then we review our selection to ensure the devices can never be damaged by excess input signals that might demand massively higher drain currents.
Click to expand...

Hi Andrew,

This is a good approach. Of course, Idss, even for a same part number, can be different over a wide range (unless the devices are available binned to Idss ranges). It is also desirable that in this design the P and N Idss values not be hugely different (probably well less than a factor of 2 apart). Also, of course, if using discretes for the pairs, it is important that their Vgs at the chosen operating point be matched.

I usually run the devices at Idss of 50% or less of Idss. Not a huge benefit in noise to running the JFETs at higher tail current, since JFET input voltage noise in nV/rt Hz only goes down as the fourth root of operating current.

Cheers,
Bob
 
#1,348
First teaser for DH-120C

Here is the first attempt and teaser of a DH-120 (PC-25) board, based on Bob Cordells design.

I will put it back in my desk, and in about 2 weeks i will go and review it again or cancel it, and start new again.

BR
Guenther🙂
 

Attachments

  • DH-120C.jpg
    DH-120C.jpg
    138 KB · Views: 619
#1,350
mlloyd1 said:
cool
🙂
a little hard these on old eyes.
i'm curious - looks like you're using a pair of singles vs. monolithic duals for the mirrors?

mlloyd1
Click to expand...

Thanks,

as i wrote it's only a check if all device fit on the original board size of the DH-120. And it looks good for my 2 DH-120 to get tuned up, that i can still modify the placements.

For the final version it's planned to use the monolithic duals, but there i am struggling with two points.

a) i must create a symbol of the SOT-23-6L and/or the SO-8L version of the dual monolithics for my PCB layout program.

b) the availability of the LSK489 and the LSJ689 here in Germany. The local distributor asked for €120 for 5 of each device. So i must find someone in the States, who will buy these for me at Trensetter and send it to me (prize in US is about $60 +S&H).

The major plan for me is building a DH-220C with onboard 2x2SK1058 and 2x2SJ162 , since i have 10 genuine Hitachi ( no Renesas) pairs left over.

BR
Guenther
 
#1,353
schillg11 said:
Thanks,

as i wrote it's only a check if all device fit on the original board size of the DH-120. And it looks good for my 2 DH-120 to get tuned up, that i can still modify the placements.

For the final version it's planned to use the monolithic duals, but there i am struggling with two points.

a) i must create a symbol of the SOT-23-6L and/or the SO-8L version of the dual monolithics for my PCB layout program.

b) the availability of the LSK489 and the LSJ689 here in Germany. The local distributor asked for €120 for 5 of each device. So i must find someone in the States, who will buy these for me at Trensetter and send it to me (prize in US is about $60 +S&H).

The major plan for me is building a DH-220C with onboard 2x2SK1058 and 2x2SJ162 , since i have 10 genuine Hitachi ( no Renesas) pairs left over.

BR
Guenther
Click to expand...

Here are a couple of things worth noting that I did in my design:

1. I did not confine myself to the original board size in the DH220; I overlapped the board at the rear end of the heat sink. In that area is where I put the input stage.

2. I just used discretes for the current mirrors; they have significant degeneration.

3. I mounted the VAS and driver transistors back-to-back, each pair, on small transistor heat sinks.

Try contacting Paul Norton at Linear Systems to see if you can get some LSK489/LSJ689 samples.

Cheers,
Bob
 
#1,357
the next step

hi

this is my version of a DH-220C with 2SK1058 and 2SJ162 power stage. The power transistors will be mounted under the PCB and the PCB will be fixed on the heatsink

Bob Cordells design modified for my use.

BR
Günni
 

Attachments

  • GS_DH220C.jpg
    GS_DH220C.jpg
    177.3 KB · Views: 567
#1,359
philnext said:
I would like this diagram of circuit for my DH-200 thanks
Click to expand...

hello philnext,

if you want the schematics, i think that is no problem. The layout of the pcb is not made for replacement of the existing PC-6 (DH-200) or PC-19 DH-220 and DH-500) from Hafler.
It's a design with already integrated power transistor, which are not seperately mounted on the heatsink as in the DH-200. I also replaced the 2SK134 and the 2SJ49 (TO-3 cases) by the 2SK1058 and the 2SJ162, which have the TO-3P cases (totally different mechanics).

Best regards
Günni
 
#1,360
schematics

Thank you I could use the schematics I have no problem bread boarding them on a RS pro boards. I do believe in the pwr supply mod for the main driver board if that's what we are talking about. I already did some mods but I need one that may improve the mid range on amp. I also found two confusing bias settings that I don't know if that will make an improvement. I set mine for 250 ma each side, though I have seen a higher value for the Dh-200 of 275 ma so what is the correct one and will this setting alone change sound quality? I drive mine with a mini tube amp that makes it sound better I think. More questions: what is type of Shockley diode do I buy I guess 2, voltage and current rating. So what is the main mods for sonic quality improvement. Thanks again. Phil
 
You must log in or register to reply here.
Top Bottom