Hi slowhands,
I'll just ask around up here. I saw a picture that looks the same as a 2500. It's a mystery why I hadn't heard of it before since I'm in the Toronto area. There was an awful lot of Marantz sold around here.
-Chris
I'll just ask around up here. I saw a picture that looks the same as a 2500. It's a mystery why I hadn't heard of it before since I'm in the Toronto area. There was an awful lot of Marantz sold around here.
-Chris
I think the last 2600 on eBay sold for over $3500 IIRC...
slowhands...I find the 1250 very easy to work on. I've rebuilt/repaired many of them. I have 5 of them here right now to be looked at...
The 1980 is not quite so simple, but I'd rather work on it than most Marantz's. 😉
djk...I looked at the schematic for the Leach, and didn't see what you were referring to...can you post a schematic of that section, or a link to where I can see it?
slowhands...I find the 1250 very easy to work on. I've rebuilt/repaired many of them. I have 5 of them here right now to be looked at...
The 1980 is not quite so simple, but I'd rather work on it than most Marantz's. 😉
djk...I looked at the schematic for the Leach, and didn't see what you were referring to...can you post a schematic of that section, or a link to where I can see it?
Leach uses four discrete diodes hooked in the base-collector circuit as shown.
For the Pioneer the D2 stack would be replaced by a single resistor composite of R25, 26, and the VR2 would replace the composite of P1, R27.
The base of the Vbe transistor would then connct to the node of D3, VR2, the collector to the node of D5, D2, and the emitter to the node of VR2, D6.
The Vbe does not go on the heatsink, only the existing D3 stack. The added Vbe transistor should be easy to tack in without even needing a piece of perf board. The BE just tack across VR2, and a short jumper from the C to the D5, D2 node.
Great setup, but too complex. If this were my own receiver, then maybe I'd fool with it. While this IS a learning experiance for me, I've already got a bazillion hours into resurrecting this one just to get to the point it is at now. Like I said, my goal is to keep the changes simple, and as minimal as possible to make this an easily repeatable procedure.
I've got a 'beater' 1980 that I will be doing all of this to sometime in the future, and on that one I have no problem with complexity, but this first one needs to follow the KISS rule.
I've got a 'beater' 1980 that I will be doing all of this to sometime in the future, and on that one I have no problem with complexity, but this first one needs to follow the KISS rule.
I need to be able to help someone through email or on the phone to get this modification done. No, it is NOT too complex for me, but have you tried walking someone through a repair who basically knows just enough to be dangerous?djk said:
This needs to be simple so that the chances of success for the semi-clueless is as high as possible.I like the idea, and when I dust mine off and get to replacing the transistors on it, I'll do it. But for this first one, simplicity is king.
To be perfectly honest, I don't know why Leach bothered with such a complication. Keeping the capacitor on-board with the driver stage, and moving just the Vbe multiplier transistor off-board (with a small C across C-B) would have brought most potential noise-injection problems to the same level as having the diode string only off-board. Also, tacking a TO-126 transistor onto a heatsink is a LOT easyer than doing the same with 4 diodes, that were never intended for that sort of thing.
Granted, both approaches must have reliable conenctors - loss of contact generates 'infinite' bias and output transistor death.
@EchoWars: Croatia is indeed beautiful, it is rare to be able to offer everything from mountain to mediterranean and continental climate, landscapes, cuisine and mentality within a couple hour's drive. Some parts of it truly are close to paradise... it would be nice if a bit more people here realized it, and took better care of it, though 🙁
Granted, both approaches must have reliable conenctors - loss of contact generates 'infinite' bias and output transistor death.
@EchoWars: Croatia is indeed beautiful, it is rare to be able to offer everything from mountain to mediterranean and continental climate, landscapes, cuisine and mentality within a couple hour's drive. Some parts of it truly are close to paradise... it would be nice if a bit more people here realized it, and took better care of it, though 🙁
EchoWars said:
Hey Echo!!
I believe i have a handfull of the org. types, purchased for some Luxman R1500 fixing a lot of time ago!!
Mail me at support@lcaudio.dk next time you see this, with the types again, then i'll look it up in the drawer...
Lars / Viborg - Tech Support for LC Audio, www.lcaudio.com
Thanks for looking nillerviborg...it was worth a shot. 🙂
Finally got my bench cleared, and the 1980 put back on for a few tests. To reduce the emitter resistor values, I simply clipped some test leads to the existing resistors, and put some others in parallel to achieve the effective value I wanted.
Anyway, turns out that for every 10 ohms of reduction of the driver emitter resistor value, I get 'about' a 0.5mA reduction in bias current through the outputs. By reducing the driver emitters from 100 ohms to 75 ohms, bias current is reduced from about 35mA to 32mA. I don't think this is worth the trouble of changing the resistor, and I'll leave the stock 100 ohm resistors in place.
On the positive side, the bias at least does stay lower than what I'd like to set it to...stablizing at about 35mA. I need to complete the second channel with the new transistors, and see if it behaves in a similar fashion. If so, then I might simply reduce the size of the bias adjust pot from 100 ohm to 50 ohm, set the bias to about 60mA, and call it good.
Once I get around to trying all this on my own dead 1980, I shall very possibly give djk's Vbe multiplier a try...
Is this what you had in mind?
Finally got my bench cleared, and the 1980 put back on for a few tests. To reduce the emitter resistor values, I simply clipped some test leads to the existing resistors, and put some others in parallel to achieve the effective value I wanted.
Anyway, turns out that for every 10 ohms of reduction of the driver emitter resistor value, I get 'about' a 0.5mA reduction in bias current through the outputs. By reducing the driver emitters from 100 ohms to 75 ohms, bias current is reduced from about 35mA to 32mA. I don't think this is worth the trouble of changing the resistor, and I'll leave the stock 100 ohm resistors in place.
On the positive side, the bias at least does stay lower than what I'd like to set it to...stablizing at about 35mA. I need to complete the second channel with the new transistors, and see if it behaves in a similar fashion. If so, then I might simply reduce the size of the bias adjust pot from 100 ohm to 50 ohm, set the bias to about 60mA, and call it good.
Once I get around to trying all this on my own dead 1980, I shall very possibly give djk's Vbe multiplier a try...
Is this what you had in mind?
An externally hosted image should be here but it was not working when we last tested it.
Hi EchoWars,
The values may change when you experiment. I'd put a resistor in series with the (now) 5K pot to set the max current. The Vbe multiplier would normally go from the cathode for D5 to the anode of D6. This replaces D2, D3 and VR2 completely.
-Chris
The values may change when you experiment. I'd put a resistor in series with the (now) 5K pot to set the max current. The Vbe multiplier would normally go from the cathode for D5 to the anode of D6. This replaces D2, D3 and VR2 completely.
-Chris
If I did that the transistor would have to be heatsink mounted. The D3 diode is already there, and I'm relying on it for bias compensation.
How about this idea:
Set your bias for 45 mA Ic
let it sit for an hour and check
if it is still OK then run it at half rated power and check the Ic every minute for the first 10 minutes or so.(removing signal source when measuring) Then increase test time and time between measurements.
If the diodes still hold Ic to a reasonabe level then you are done.
Set your bias for 45 mA Ic
let it sit for an hour and check
if it is still OK then run it at half rated power and check the Ic every minute for the first 10 minutes or so.(removing signal source when measuring) Then increase test time and time between measurements.
If the diodes still hold Ic to a reasonabe level then you are done.
Hi Doug,
They will hold the bias to a constant semi-constant level. It was a successful commercial product.
EchoWars would like to change the bias beyond the normal adjustment range. Also, sometimes these special diodes are found broken and they can not be easily replaced. A Vbe multiplier is then the best option.
-Chris
They will hold the bias to a constant semi-constant level. It was a successful commercial product.
EchoWars would like to change the bias beyond the normal adjustment range. Also, sometimes these special diodes are found broken and they can not be easily replaced. A Vbe multiplier is then the best option.
-Chris
complicated idea:
put in a isolated regulated suppply and a small reverse voltage to reduce the effective potential of the compensation stack.
🙂
put in a isolated regulated suppply and a small reverse voltage to reduce the effective potential of the compensation stack.
🙂
If you change it to be exactly the same as the Leach, it will work exactly the same as a Leach, because both amps have tripple EF output stages, and the four diode stack on the heatsink.
I gave the reference numbers from the Pioneer schematic you provided, and the reference numbers from the Leach schematic, and gave you a step-by-step connection to adapt the Leach design to the Pioneer layout.
I gave the reference numbers from the Pioneer schematic you provided, and the reference numbers from the Leach schematic, and gave you a step-by-step connection to adapt the Leach design to the Pioneer layout.
- Status
- Not open for further replies.