diy81 Progress

[ToroRojo]

I replaced the current limiting resistors with 2.4k. They still burnt out.

The specs for the leds I have say they have an If (max) of 30ma. Strange (not to mention frustrating) that they should burn out so easily at 10ma.

 

[ToroRojo]

I finished wiring up all of the eq switches. And I figured out why the +15 line position didn't work (it actually did work, I just didn't have the knob lined up right. . ).

The unit works great when the eq is not engaged. When the eq is engaged, after about 2 minutes of operation, it starts distorting. After another minute, the output starts to drop out in a somewhat periodic fashion. It is possible to get oscillation by playing around with the eq switches. It is really easy to get it to oscillate by cutting frequencies. There is also a considerable output boost when the eq is engaged.

I triple checked all of my wiring. Everything appears ok. I guess its time to bust ot the oscope.

 

[Tekay]

Check tantalium and electrolytic polarity in the signal pathand the components in the input bias stage on each 338. It's easy to follow the signal through the EQ section, just check the signal after each stages output cap.

 

[ToroRojo]

Thanks for the hints Tekay.

I figured out the oscillation problem. It's a total bonehead move, but I left out the 68pF capacitor that mounts on the bottom of the 338 boards. That's fixed now.

The distortion and drop outs after 2 minutes of use is still present though. If I don't input a signal, but leave the eq engaged, there is a periodic signal present in the amp. It is a low frequency ba-bump at (very roughly) 50bpm. I can eq it! I'll dig into it some more tomorrow...

 

[ToroRojo]

I traced the distortion and drop outs to the lower presence section. The fact that the problem occurs after warming up (and continues to get worse the longer the unit is on) makes me think that one of the transistors on the 338 board in that section is faulty. I'm going to build another 338 and put it in there and see if it clears up.

 

[ToroRojo]

I built a new 338, and replaced the suspect card, but the problem persisted, so I started checking capacitor polarities.

Sure enough, there is capacitor and resistor in series between ground and pin 7 of the low-mid 338 card. I had it set up as follows:

pin7 -> 3k6 - > 22uF+ -> gnd

it should be

pin7 -> +22uF -> 3k6 -> gnd

I had it connected as shown in the artwork on the board, but that is incorrect. It looks like Frank's latest revision of the component placement has this corrected.

The only problem I have with the unit now is the fact that the output goes up by roughly 6db when the eq is engaged. I've narrowed th eboost down to the high-mid eq section, but I haven't figured out why it happens.

 

[ToroRojo]

Figured it out. R30 on the board artwork needs to be 30k. I had a 15k in there.

Just need to get the leds to last more than 5 minutes and this thing is golden!

 

[AnalogPackrat]

:thumb: If you can describe briefly how the LEDs are wired it might help (or point us to a schemo/board layout). Do the current limiting resistors go to +24 or somewhere else? How are they switched? The only thing I can think of is that there could be a current spike at turn on that's killing the things since you seem to have a good sized limiting R in there.

A P

 

[ToroRojo]

Quick MS Paint mockup:

Any help is greatly appreciated. :guinness:

 

[AnalogPackrat]

Well, there you go... There is no limiting resistor between the cap and the LED, so when you first switch it on, the cap spikes a large current through the LED. A few spikes and she's a DED (dark emitting diode), Jim. The limiting resistor needs to go between the LED and the cap.

<edit> More analysis: When the switch is open, the cap is sitting at 24V. When you close the switch, the cap discharges until it is down to Vf of the diode (about 1.5V for red, higher for other colors). During this discharge, a large current spike flows through the LED because there is no limiting resistor between the stored charge in the cap and the LED.

A P

<edit> Additional info: You should have a current limiting resistor for each LED. Otherwise any I-V curve differences will cause one to hog current and they won't all work (or their brightness will vary depending on how many are on). 2k-3k would probably be fine.

 

[ToroRojo]

Thanks AP! I replaced the resistor in the circuit with a 100R, then moved the 2k4 to be in series with the led (with both being in parallel with the cap). It's working now! The circuit I put up is per led, fyi. I guess the original 1081s used leds with built in current limiting resistors?

Anyways, here's some photos!

 

[peter purpose]

>>I guess the original 1081s used leds with built in current limiting resistors? <<

Lamps.

 

[Kid Squid]

Hiya Toro,

Looks the Absolute buisiness Fella, :grin:

Steve :thumb:

 

[t-wurst]

absolutely breathtaking :thumb:

 

[kiira]

That is absolutely gorgeous.

BTW ar ethese the boards folks are using? This guy has 7 of them for sale.

http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&rd=1&item=260048189615

Kiira

 

[Kid Squid]

Kii,

those are the same boards, but take into consideration of the cost of the iron - mic, line and o/p, as well as the concentric switches / gain switch.

those damn concentric switches............
I need another three, but can't get them anywhere ......yet :wink:

Regards,
Steve :thumb:

 

[ToroRojo]

Thanks guys. Since I finished the first one, I built a second one!

Here's a picture of them in action (with some other gear). I usually keep 'em racked up, but I was travelling and had to travel as light as possible. I don't like to ship or check gear...

 

[Guest]

[quote author="Kid Squid"]
those damn concentric switches............
I need another three, but can't get them anywhere ......[/quote]
I've got them but not the pots with the mega long spindles

 

[Kid Squid]

UK,

what diameter are the pot spindles you have now ?
do you want me to see if i can turn a few of the spindle ' converters' out in work,
I'm going to need a few myself !
what pots are you using ?


Toro - Just to rub it in

Thanks guys. Since I finished the first one, I built a second one!

:wink:

Where did you get the cases and the frontplates from, I NEED two sets,
:roll:
Be mucho appreciated if you could let me know Old bean.

Best Regards,
Steve :thumb:
Regards,

 

[ToroRojo]

KidSquid -

I half salvaged, half made the cases. I purchased some old neve modules that someone had gutted for the 283 cards and trannies and used what I could from the metalwork which was pretty much just the side rails. I fabricated everything else with a sheet metal brake, some tin snips, a dremel, a drill press, and some pliers.

I purchased the front panels from a place in the UK. Pro audio design? I can't remember...

- Old bean :grin: