Roender's FC-100 prototype and builder's thread

[atupi]

Work in progress ...

I want to show you some pictures of my work.
This pcb's are just lovely and my soldering iron is very happy with them
Still missing some parts but hope to have it soon.

 

[atupi]

I intend to buy some cheap heatsinks for the first tests before i sped some big $$$ on some fancy case from modushop.
Would be this heatsink ok for preliminary tests: Radiator aluminiu - 160x110x32 mm-1322 - Okazii (121721113) ? (160x110x32 mm)

Adrian

 

[AndrewT]

It will run very hot. I guess Ts>>50°C and possibly >70°C

 

[atupi]

Andrew for wich power output you made that calculations ? I was having in mind not running at full power, maybe 20-30W.

 

[AndrewT]

It's not power output that is the problem.
It's quiescent power dissipation.

At normal running condition where average power output is <1% (<-20dB ref max) of maximum power output you will find that the heatsink temperature does not appear to change whether delivering that ½W or not.

 

[Rudi_Ratlos]

Adrian, the power dissipated by a FC-100 channel in idle status alone is about 35W!
An adequate heatsink should have a performance of about 0.5° C/W.

Do you have a small fan to assist cooling during the build-phase? Then you can use the heatsink shown.

Best regards - Rudi_Ratlos

 

[atupi]

Thanks for your advice Rudi!
I think i can find some old computer cooler to use.

 

[atupi]

Finally i've received 2SA/2SC transistors and as with MJE's is impossible to find PNP-NPN matched pairs but i have an idea and i ask you to tell me if i'm wrong or not.
In the attached picture i see two Sziklai pairs and i was thinking to "match" the equivalent beta of the Sziklai compounds.
Is my thinking correct ?

An externally hosted image should be here but it was not working when we last tested it.

 

[franzm]

At dc it works perfect. For ac you should simulate the circuit and see what effect the imbalance of the corresponding transistors has on distortion.

 

[Telstar]

Change the models with the measured beta and try, I think is a good idea.

 

[atupi]

Usually PNP transistors have lower hfe than their NPN pairs but i have found with my DMM that this rule is not apply for MJE 15034/15035 and is not possible not match the total hfe of Sziklai transistors (NJE/2SA/2SC pairs).
Is that normal for this MJE transistors to have bigger hfe for pnp or should i measure their hfe in another environment than my DMM ?
This matching is my last issue in finishing this project

PS. hope my english is understandable this time

 

[Telstar]

Usually PNP transistors have lower hfe than their NPN pairs but i have found with my DMM that this rule is not apply for MJE 15034/15035...

No, you're not alone.
Modern PNP transistors can indeed have higher hfe and be more linear!

 

[PeterLeest]

These are the trace curve for my MJE 15034/15035 & NJL0281/3020. It looks like the NPN TRs are not so linear as compared to the PNP ones. Also, Hfe gaps are quite huge.

 

[AndrewT]

I have measured quite a few ONsemi devices at various spot Ic values.

I have never found the big variations in hFE (once above 1% of Icmax) that are reported in post713.

 

[Rudi_Ratlos]

Andrew: it depends.
I have measured more than 300 pieces of NJL0281G/NJL0302G in the last year, always using the "matching circuit".

Depending on the production date / date of my order / company where I ordered: the transistors either match very well (in Claus1968' case)
or not so well (in Klaus' case).

Best regards - Rudi_Ratlos

 

[AndrewT]

Rudi,
I am not referring to matching.
I am referring to post713 showing the variation in hFE for a single device as Ic is varied.
For that specific test situation I have NOT found Onsemi devices to have significant hFE variation (non linear) as Ic is varied, once the Ic value is above a sensible minimum value.

 

[Rudi_Ratlos]

Andrew: I understand. I am with you then.

Rudi

 

[atupi]

Wich to choose ?

I have the following transistors on hand, please advise me wich pairs to choose for Q1,Q3,Q19,Q20 ?

An externally hosted image should be here but it was not working when we last tested it.

An externally hosted image should be here but it was not working when we last tested it.

 

[AndriyOL]

Does anybody have an idea what could be a reason of this amplifier topology starting to short at once from 5.6mV and higher VRe or from toching by test leads as a trigger (when VRe close to 5.6mV) or by itself after 5-10 minutes? During the short condition VRe shows 33mV, Vbe-s are in range.

 

[PeterLeest]

Rudi,
I am not referring to matching.
I am referring to post713 showing the variation in hFE for a single device as Ic is varied.
For that specific test situation I have NOT found Onsemi devices to have significant hFE variation (non linear) as Ic is varied, once the Ic value is above a sensible minimum value.

The question TR MJE15304 is procured from RS component (RS Stock No. 463-098P.

I did three consecutively test on same transistor. The Hfe would increase a bit in each test (is it affected by heat ?). Also, for 15304, the increase seems higher than other transistors, eg. MJE15305, NJL0302/2801. all 5 TRs I puchased behave same characteristic.

could it be a fake?