DIY JE-990 DOA

[evm1024]

4 passes through the core in the JH990C that I took apart. The photo shows it all....

 

[structuredloud]

Just finished building the ptown kit. Pulled my Hardy out and put this sucker in. Sounds almost the same! I put 33k in R10+11 so it would fit my 2520 type circuit. Great fun, especially the elaborate diode transister heatshrink orgy. I can't imagine how the Hardy Co. manages to put these together for $50, what a deal (or very handy robots).

 

[3nity]

Guys lacker doesn't melt with the heat? providing enough contact??

 

[Ptownkid]

you should scrape the wire with a knife...and once in place, check for continuity.

 

[3nity]

I should have done it yeah. i figure this out too late.... they soldered to the boards on my 1st and on the 2nd they are fine.... i scraped before solder it to the board!!

 

[fazeka]

I have some questions on the Ptownkid's 990 Op Amp kit. I’ve never put a discrete opamp together before and I am not sure on a couple of things. Yes, I've been reading the groupdiy archives. Still a little confused, so maybe y’all can shed some light.

OK, first: the diodes that thermally couple to the output transistors. I am not exactly sure what it means to "thermally couple" something but I suspect I need to contact one of the legs of the output transistors with the body of the diode? If this is correct, is this in lieu of a heat sink and if so, then why not just use a heat sink? Also, does it matter which leg of the transistor is coupled to the body of the diode?

Secondly, regarding the inductors. Looking at some threads on the groupdiy forum, I see that there are two ways of threading the ferrite core.

http://www.groupdiy.com/index.php?topic=18409&postdays=0&postorder=asc&start=30

From the top of mnats’ inductors in his 990 pic, it looks like he did it the first way as in the thread, but some say the method as seen in Jensen's patent gets closer to 20uH. However, looking at the PCB from Ptownkid's kit, the overlay implies yet a *third* method of wiring the core as the entering and exiting holes from the core into the PCB are different than that referenced in the aforementioned thread (!). Not sure if I am misinterpreting or agonizing over nothing?

Lastly, I noticed that transistors Q3 and "CR3" (seems to me that is a typo) are facing each other. I guess these are "thermally coupled", as well? If so, do I need to dress the "faces" of these transistors before coupling them with thermal grease or compound of some sort? Also, is that a rubber band fastening them together?

Your help is appreciated, sorry for all the questions. Unfortunately, I am still an op-amp newbie...

 

[Ptownkid]

For the diodes, refer to this pic

As for the winding pattern, I'm not sure what the consensus was, the results are contained in one of the threads somewhere, but I used mnats pattern.

edit, found the thread http://www.groupdiy.com/index.php?topic=18409&postdays=0&postorder=asc&start=0

The two transistors that face each other should have thermal compound between them, and the elastic band looking thing is a piece of heat shrink.

Hope that helps

 

[Adam8ch]

Got the ptownkid kits and was about to sit down and solder them, but I realized I had a question about the Q3/CR3 and CR9/10 thermal coupling:

Is it sufficient to meet the diodes and transistors touching (maybe with thermal compound between them), or is it necessary to use an epoxy or adhesive (as in dbranco's pics and evm1024's post). The pic that Ptownkid said to reference just looks like there's contact but nothing definitively bonding the the diode and the Q legs together. If anyone could shed some light on this subject, I'd be much appreciative.

Sorry for the noob question, but I can't get a definitive answer from the threads (this or otherwise). Just want to make sure I have a clear idea of what I'm doing before I do it (so any further mistakes are just my handiwork and not my judgement)! Thanks

 

[b3groover]

From what I've read and seen, most folks use a thermal compound between them and then a small piece of heat-shrink to hold them together.

I'm about to start building mine. I've been reading as much as I can before I start. Gonna plop them in some MAP preamp cards and see what they do!

 

[MartyMart]

Is it true that the 990 doesn't like Hi Z input ?
Was just checking mine out with a simple circuit ( not balanced/no traffo ) from guitar in
and it seems to be overdriven and volume just drops during pot movement !
I guess that for testing, I can't treat it as a "normal" 741 for instance ??
( running +/- 24v of course, nothing burned up & voltages look OK )
Marty.

 

[3nity]

sad but true.

 

[delaymix]

[quote author="MartyMart"]Is it true that the 990 doesn't like Hi Z input ?
Was just checking mine out with a simple circuit ( not balanced/no traffo ) from guitar in
and it seems to be overdriven and volume just drops during pot movement !
I guess that for testing, I can't treat it as a "normal" 741 for instance ??
( running +/- 24v of course, nothing burned up & voltages look OK )
Marty.[/quote]

Can you or someone explain reasons why this happens?

 

[burdij]

You might want to view the 990C datasheet on the John Hardy site and make sure your application is configured correctly. This amp should operate in any circuit where a monolythic op amp can be used, except for those requiring extremely high input impedances. The input impedance of the 990C (non-inverting input) is > 10Meg. Ohms.

http://www.johnhardyco.com/pdf/990-2007.pdf

 

[keefaz]

On this datasheet, at the MPC1 shematic near the end, I see the 990 is powered with unregulated voltage (just filtered), refering to the power supply section... I thought it needed regulated voltage as other opamps :?:
I have to try this

 

[Samuel Groner]

Is it true that the 990 doesn't like Hi Z input?

He does and he doesn't. They surely will pass audio for any proper implementation (there's something else wrong with your circuit--have you considered gain and input bias currents?) but noise will not be optimum--JFETs are much better at high source impedances.

I see the 990 is powered with unregulated voltage.

What makes you feel that the +/- 24 V should be unregulated? If you like your 990s I would not feed them with unregulated voltage--in case it gets well above 48 V it will die, die, die...

Samuel

 

[MartyMart]

Yes I think I ended up with just a badly biased little circuit !
Started off as a simple gtr "booster" and ended up as a JH mic pre without
any traffo up front !!
Will turn it into either a "stock" mic-pre with traffo or the JE16 perhaps.
It's intended to end up as an API 312 with Cinemag/4804.
MM.

 

[keefaz]

[quote author="Samuel Groner"]
What makes you feel that the +/- 24 V should be unregulated? If you like your 990s I would not feed them with unregulated voltage--in case it gets well above 48 V it will die, die, die... [/quote]
Hey thanks for the correction, I just re-checked the shematic and I made a confusion, I thought on a quick read that the diodes at power supply input were a bridge rectifier...

 

[rodabod]

[quote author="evm1024"]The wire in John Hardy inductors is not insulated. Just 28 gauge tinned copper.[/quote]

I'm just figuring out how to make my inductors. Is it ok to use uninsulated copper wire or does it need to be enamel-coated? The inductor core is slightly conductive, but I wasn't sure if this could be a problem.

If I do need enamelled wire, does anyone know any common sources other than buying magnet wire from suppliers - it's quite expensive.

 

[rodabod]

Wait a second, jus found an old inductor in our spares cupboard which I will butcher.

 

[skipwave]

[quote author="burdij"]You might want to view the 990C datasheet on the John Hardy site ...
http://www.johnhardyco.com/pdf/990-2007.pdf[/quote]

Something occurred to me whilst looking at that JH paper. On page 4 under Application Notes, Figure 1 shows "Traditional Mic Preamp" config. However, with C3 between ground and RV1 (gain set pot), isn't this current feedback mode? Yet, there is a bandwidth limiting cap, C2, across the feedback resistor, R3.

I thought the phase lag introduced by C2 could cause instability. Although, my only experience to date with current feedback design is the Tamas FETBlokes (which sound phenomenal). The JH paper got me wondering if there is some benefit to running DIY990s in current feedback mode.