Hi, I hope to get some help here.
I wanted to re bias my CREEK evo integrated amplifier after it had been dropped in transit
The dc mv drop over both .22ohm 5 watt emitter resistors was about 1mv each!. I re adjusted that to about 7.5mv. The right channel was adjusted to be the same .
The trim pots I used for this were the ones closer to the heatsink.
Dc offset was below 8mv for both channels after 20 minutes at idle
My question is about the two trimpots farther away from the heatsink,what do these do? I did give them a turn one rotation in both directions thinking they might be for the dc offset but it didn't seem to affect that measurement.
Can anyone tell me what these trimpots do and what I need to measure and adjust with them.
Cheers
Chilli
I wanted to re bias my CREEK evo integrated amplifier after it had been dropped in transit
The dc mv drop over both .22ohm 5 watt emitter resistors was about 1mv each!. I re adjusted that to about 7.5mv. The right channel was adjusted to be the same .
The trim pots I used for this were the ones closer to the heatsink.
Dc offset was below 8mv for both channels after 20 minutes at idle
My question is about the two trimpots farther away from the heatsink,what do these do? I did give them a turn one rotation in both directions thinking they might be for the dc offset but it didn't seem to affect that measurement.
Can anyone tell me what these trimpots do and what I need to measure and adjust with them.
Cheers
Chilli
Attachments
If the bias level dropped to 1mV, it wasn't due to the setting of those multi-turn pots. There must be a fault elsewhere for the bias to mysteriously wind up at 1mV on both channels and yet a simple adjustment restores it. That seems suspicious to me.
With no free schematic available to us unless we are Creek service agents, it's not possible to be 100% certain what the second set of pots are intended for. It's most likely they do adjust DC offset voltage, yet again there is either the same problem in both channels (very unlikely) or your measurements are not being performed correctly. <8mV suggests you can't be certain but to be honest, it doesn't matter when any stable offset below ~40mV is just academic.
Usually, there are output relays fitted to all newer power amplifiers (I think I see one behind the heatsink in your pic.) and this should click as the relays close after a short delay on power-up, connecting the output (+) terminals the amplifiers. Only then will you be able to measure DC offset at the terminals. Otherwise, you could locate the output rails before the relays and check the + voltage there if the protection circuitry has shut down the relay driver. Check that your measurement is for real.
With no free schematic available to us unless we are Creek service agents, it's not possible to be 100% certain what the second set of pots are intended for. It's most likely they do adjust DC offset voltage, yet again there is either the same problem in both channels (very unlikely) or your measurements are not being performed correctly. <8mV suggests you can't be certain but to be honest, it doesn't matter when any stable offset below ~40mV is just academic.
Usually, there are output relays fitted to all newer power amplifiers (I think I see one behind the heatsink in your pic.) and this should click as the relays close after a short delay on power-up, connecting the output (+) terminals the amplifiers. Only then will you be able to measure DC offset at the terminals. Otherwise, you could locate the output rails before the relays and check the + voltage there if the protection circuitry has shut down the relay driver. Check that your measurement is for real.
Thanks Ian, I will try the things you suggested tonight.Maybe I did not wait long enough for the dc offset to settle after I played with the rear pots
You said the amp. had been idling 20mins and DC won't change much with the cover off. Still, it should remain relatively stable because it is determined by the cool input stage, not the hot output stage. The issue is about whether the output relay was closed when you made the DC offset measurements. If not, the reading will be nonsense.
Well thanks for your advice Ian.Those rear pots ARE for the dc offset, I measured from the output coil before the relay and got both sides easily adjust to around 1mv and stable.
Also I must have measured wrong last night as when I measured tonight between the output emitters on each channel there was around 0mv?...anyway when I adjusted them this time to about 16mv ( across both .22R resistors ) I could finally feel the heatsinks get toasty warm ( not hot ) for the first time since I have owned the amp.
This is also my first attempt at biasing an amplifier ( its probably obvious )
Thanks for your help 🙂
p.s It is sounding great, sweet and dimensional
Good that it's sounding great 🙂 I'm interested to know, since it was the reason for your thread, if there were signs of damage due to being dropped, such as bent steel case, misaligned or broken PCB mountings, connectors etc. I can't see anything from the pics but even though yours are are excellent quality, pics can be deceptive.
Then there's the question of bias setting. How did you determine that the idle current was to be set to 8mV across 0.22R or 36 mA ? It's a low current for an EF output stage design but seems plausible for higher grade Sanken output transistors. However, did that value come from a service manual or someone's educated guess?
Then there's the question of bias setting. How did you determine that the idle current was to be set to 8mV across 0.22R or 36 mA ? It's a low current for an EF output stage design but seems plausible for higher grade Sanken output transistors. However, did that value come from a service manual or someone's educated guess?
I have only had the creek for about a month. I imported it from the Uk and the box had some damage to it although the amplifier itself was unmarked. So I was concerned that although the amplifier sounded ok it was generating no heat.
I have included a couple of schematics of older creeks that show about 6 mv across the .22 ohm 5 watt resistors.
Is it possible that my amp being bipolar output compared to the earlier mosfet outputs would require less biasing? or is this a non issue.
The reason I mention this is that I have just received an email from Mike Creek and he said there should only be around 1.5mv across each .22 ohm resistor!
His reply was
"To reduce crossover distortion to a reasonable level, the voltage drop across each 0.22Ohm resistor should be approximately 1.5mV, or 3mV across both. Using more bias than that will cause the heatsink temperature to be too high. The second pre-set is for small DC adjustments."
Attachments
Last edited:
I see. It's worth saying that mosfet and bipolar devices usually have quite different bias requirements depending on the type of device, the actual chip technology and the configuration of the output stage. A typical class AB CFP design bipolar amp. could have a low bias of perhaps only 10-20mA but a mosfet design could require 100-200 mA per pair of output devices before the crossover distortion disappeared completely and the sound became sweet and clean.
You can't predict the electrical design and bias requirements by the brand name, as the title "Evo" might suggest. Creek started with bipolar and have changed technologies with each designer's influence. As this is a bipolar design and you find the sound acceptable with 36 mA, it could be a fortuitous outcome, provided this current level remained stable when the amplifier was then played at solid volume levels with the cover in position and everything became quite warm yet bias measured as fairly stable. It's probably a safe bet but test again after a listening session - bias stability is critical to the longevity of your amp. You certainly don't want thermal runaway and a smoking amp showing you why you need to be careful with bias.
However, the design setting according to your attached setting procedure is a very low 1.5mV across 0.22R - only a tiny 7 mA and whilst that will kill most crossover distortion from a CFP output stage, it's going to sound a tad sharp for my liking and perhaps fussy with speakers regarding the remaining distortion artefacts.
36 mA is obviously a lot higher than 7 mA and explains why you thought the bias was zero and needed adjustment because it felt cold - unlike the majority of amplifiers which use emitter follower designs amplifiers with 50-100 mA bias current. I would back your setting off to 15 mA or a measurement of 6 mV across both 0.22R resistors and see what you think about the sound quality and the stability of the setting. It will still run quite cool and steady now as most CFP designs do but that may be a different story in summer with 30-40C ambient temperatures. Keep a close eye on stability as summer approaches if you are going to use a higher bias setting than specified by the manufacturer. It's not an arbitrary control to adjust at will so take care and test periodically.
You can't predict the electrical design and bias requirements by the brand name, as the title "Evo" might suggest. Creek started with bipolar and have changed technologies with each designer's influence. As this is a bipolar design and you find the sound acceptable with 36 mA, it could be a fortuitous outcome, provided this current level remained stable when the amplifier was then played at solid volume levels with the cover in position and everything became quite warm yet bias measured as fairly stable. It's probably a safe bet but test again after a listening session - bias stability is critical to the longevity of your amp. You certainly don't want thermal runaway and a smoking amp showing you why you need to be careful with bias.
However, the design setting according to your attached setting procedure is a very low 1.5mV across 0.22R - only a tiny 7 mA and whilst that will kill most crossover distortion from a CFP output stage, it's going to sound a tad sharp for my liking and perhaps fussy with speakers regarding the remaining distortion artefacts.
36 mA is obviously a lot higher than 7 mA and explains why you thought the bias was zero and needed adjustment because it felt cold - unlike the majority of amplifiers which use emitter follower designs amplifiers with 50-100 mA bias current. I would back your setting off to 15 mA or a measurement of 6 mV across both 0.22R resistors and see what you think about the sound quality and the stability of the setting. It will still run quite cool and steady now as most CFP designs do but that may be a different story in summer with 30-40C ambient temperatures. Keep a close eye on stability as summer approaches if you are going to use a higher bias setting than specified by the manufacturer. It's not an arbitrary control to adjust at will so take care and test periodically.
Thanks again Ian for your support. I am away this weekend, but when I get home I will reduce the bias levels to 6mv for each channel and see how it sounds.
cheers
Paul
EVOLUTION AMPLIFIER Set-up procedure. 2009 Right Channel 1. Mid-point (DC Offset) adjustment. Use a multimeter to measure the voltage from the R-MID to Ground. Adjust R509 to measure 0V +/-5mV at room temperature (20 – 25 degrees). 2. Idle current adjustment. Use a multimeter to measure the voltage of the idle current from the R-MID to junction of R522, R516 and Q510 emitter. Adjust R532 to measure 1.5mV. +/-5% Left Channel 1. Mid-point (DC offset) adjustment. Use a multimeter to measure the voltage from the L-Mid to Ground. Adjust R609 to 0V 2. Idle current adjustment. Use a multimeter to measure the voltage of the idle current from the R-MID to junction of R522, R516 and Q510 emitter. Adjust R632 to 1.5mV. +/- 5%
There may be an error in the last sentence. It should be like this:
Use a multimeter to measure the voltage of the idle current from the L-MID to junction
of R624, R616 and Q610 emitter. Adjust R632 to measure 1.5mV. +/-5%
Use a multimeter to measure the voltage of the idle current from the L-MID to junction
of R624, R616 and Q610 emitter. Adjust R632 to measure 1.5mV. +/-5%