I find myself repeating the fact that the emitter diode has a non-linear element of resistance that decreases with increasing temperature. The resistance is inversely proportionate to the temperature on the Kelvin scale. Use this to work out where you are at. At 25 degrees C the diode resistance with have a voltage drop of 26 mV.
Another way to look at this is to ask if an aircraft is loaded to normal capacity should the length of the take off run way be shortened. That is what you do by increasing the bias.
Another way to look at this is to ask if an aircraft is loaded to normal capacity should the length of the take off run way be shortened. That is what you do by increasing the bias.
OK, you achieved to alarm me... I'll hear this amp with week-end with very attention, and with normal bias too, and I will certainly return the bias to the normal value.
But what sgrossklass wrote makes sens : "Note how the service manual suggests to "run about 5 minutes with rated output (8 ohm)" - that'll warm up the heatsink considerably. So the idea is to have it run at 2.5 mV when the main heatsink is decidedly warm / hot, not when it's just idling away".
Except if I decide to use load resistors to run my amp at full power for 5 minutes, I will never reach the point where the bias must be REALLY 2,5 mV. As I told you, the drivers heatsinks are hot, not the main heatsink, on the level I use this amp (power turned max between 9 and 10h)
It you were less authoritative in making these sort of statements and followed the service manual recommendations you would not be here preaching gospel to the converted. In post 2 PPR told you the output stages run cool AB and the drivers run in Class A.
The heat sense transistor is on the output heat sink and does not monitor the drivers so if these heat up too much you are in trouble. The collector current in these will double for every 11 degrees rise above room temperature of 25 degrees C.
In case you have not grasped my point about aircraft take off run ways this analogy refers to the driver transistors.
The bias adjustment will be a finnicky one.
Earlier I suggested you measure the voltage drop across the driver emitter resistors you do that by connecting one probe 0 volts and the other alternatively to either end of the resistor and subtracting to get the difference.
From Ohms law that will show what the current level is. If you have a temperature measurement range and probe for your multi meter use this. Don't assume you can get away without replacing the driver emitter resistors.
If any doubts or difficulties arise out of this process - ask.
Last edited:
I apologize if you had this impression, I can't be any authoritative here, I have NO technical knowledge. You and sgrossklass have points of view VERY different, if not opposite, and I just want to understand, it's only that.
Here I'm the candid, I can't show authority in anything.
When I found this amplifier, after 5 minutes playing low level the bias was around 9,5mA. So how did it not burn before I bought it ?
When you tell me "to measure the voltage drop across the driver emitter resistors", I don't know that is a driver emitter resistor. But I can measure the voltage drop across R641 and 642, if that's what it's about.
I have a basic multimeter without temperature probe. Must I measure this voltage with the resistors in place, or must I doing it after they have been changed ? (for 3 watts resistors, If I can find it) ?
Thank you for all your infos !
If you don't have a temperature gauge you can use a method for detecting either heat or cold (draughts). That is to moisten a finger to sense air movement or radiation of heat.
The former you can use to see if there is enough air circulating around the resistor bodies or too much heat coming off the driver transistors and R641 and R642 which you have correctly identified as the emitter resistors.
I see from the parts list in the service manual that these resistors which look like carbon film types are listed as being metal which puts a different complexion on my recommendation about replacing these.
The part number RD 2BJ1k appears to be a Rotel in-house identifier.
Other resistors in the list are given as metal film types and these could also be taken for carbon from appearances but in those cases I think these are original as it is unlikely these would fail in use.
I have not done any repairs on Rotel amplifiers but there are quite a few members who work in the trade possibly pngrs who could settle the question with a simple answer.
I replaced these 4 big 1 Kohms resistors today, I used 3 watts resistors instead of 2 watts.
- Resistors R643 and 644 vere dead (infinite resistance)
- After this , I measured the voltage between each end of each resistor and the ground, to get by difference the voltage across the resistor (is it a correct way ?). I got 43V for each resistor, this gave me 43mA, and 1,85 watt, for a bias adjusted around 2,5 mV.
I ordered load resistors, to put this amp in heat as described in the S. Manual for definitively adjusting this bias. While waiting for this moment, what do you think about these values ? (with 1,8 watts across these resistors, upgrading from 2 to 3 watts looks necessary). And while the main heatsink is just a little warm, the 4 aluminium for Q623 to 626 are really hot, you can't leave your hand on it.
Have a nice evening !
Last edited:
Not sure if I should start a new thread...
R641-644 (the 1K 2W emitter resistors) were all cooked... infinite resistance. I wondered why the driver heatsinks were so cool. After replacing these resistors the output bias voltage went down (it was over 10mV when I opened it up). Surprisingly, the amp still sounded pretty good even without the driver transistors biased in Class A. The main heatsink is barely warm to the touch.
I just bought a used RA-840BX3, which is almost the same as the BX4. Mine shows a drift of over 50mV. It seems to stay below 100mV. Not sure if this is temperature related or if there might be a failing electrolytic capacitor or something. Is this cause for concern?
R641-644 (the 1K 2W emitter resistors) were all cooked... infinite resistance. I wondered why the driver heatsinks were so cool. After replacing these resistors the output bias voltage went down (it was over 10mV when I opened it up). Surprisingly, the amp still sounded pretty good even without the driver transistors biased in Class A. The main heatsink is barely warm to the touch.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.