AudioGeek said:
Hello ,
I've been using cascoded DN2540 sinks for the past six months , previously just using a single DN2540 device under output stages . The difference is not too great . Never tried bipolar high-current sinks to make comparisons though . I suppose battery biased mosfets are another possibility , there are plenty of designs on Gary Pimms site to use as reference . For low mu valves like 6B4G I find it best to use variable grid bias from a negative supply and drop about 20V across the sink . This way sink dissipation and HT wastage is reduced . A centre reading meter is connected via resistors between the cathodes so the output stage can be balanced avoiding DC in the output transformer . The scheme sounds far better than any other biasing scheme used to date , I don't use anything else now .
ps Filamant transformers for the 6B4G will need an electrostatic screen otherwise the leakage capacitance will affect operation of the sink
316a
Hi,
I have been playing with differential output stages for several months now. The versions I’ve been working on use pentodes and local feedback from the plate to the grid of the output pentode.
In this configuration the output stage works as a I to V converter and has the input stage doing the duty of V to I converter. With this setup I'm getting sound quality that is really great. Super fast, clean, and dynamic range that my SE300B amp can only dream about. Frequency response of the differential amp is flat from single digit numbers to -3dB at 165Khz.
One thing to keep in mind when building/tweaking series connected (differential) output stages is that if either tube goes into cutoff the whole output stage current loop breaks. This means that you will want to run the output stage in deep class A, possibly at a higher bias current and lower plate voltage that your would used in a single ended amp. This is just the opposite of the way most P-P stages are setup. The goal is to make sure that when the output stage clips, the clipping happens to the tube approaching saturation. We want the tube approaching cut off to still have a good amount of current flowing to maintain integrity of the output stage current loop.
The amp that is currently under construction will be used to learn more about the interactions in the output stage in differential amplifiers. It will be using EL84's initially. This makes it a better choice for dinking around with than the 47P-P amp. Tubes are much easier to come by and can be operated at any angle.
There are a few CCS designs on my web page that will work for 120ma or more. The mosfets used are good for 50 watt dissipation so are much tougher than the DN2540's 15 watt rating. I also sell the boards to build the CCS's on. There are 2 new versions that have not been added to the web page yet. A new battery biased CCS that uses 27A 12V alkaline batteries that snap into battery clips instead of the lithium cells that solder into the board. And a small footprint CCS (same as the battery biased CCS's) that self-biases using the voltage reference from the Rev 5 CCS.
Here is a picture of the newest CCS's in the family
Gary
Gary P's DIY page
I have been playing with differential output stages for several months now. The versions I’ve been working on use pentodes and local feedback from the plate to the grid of the output pentode.
In this configuration the output stage works as a I to V converter and has the input stage doing the duty of V to I converter. With this setup I'm getting sound quality that is really great. Super fast, clean, and dynamic range that my SE300B amp can only dream about. Frequency response of the differential amp is flat from single digit numbers to -3dB at 165Khz.
One thing to keep in mind when building/tweaking series connected (differential) output stages is that if either tube goes into cutoff the whole output stage current loop breaks. This means that you will want to run the output stage in deep class A, possibly at a higher bias current and lower plate voltage that your would used in a single ended amp. This is just the opposite of the way most P-P stages are setup. The goal is to make sure that when the output stage clips, the clipping happens to the tube approaching saturation. We want the tube approaching cut off to still have a good amount of current flowing to maintain integrity of the output stage current loop.
The amp that is currently under construction will be used to learn more about the interactions in the output stage in differential amplifiers. It will be using EL84's initially. This makes it a better choice for dinking around with than the 47P-P amp. Tubes are much easier to come by and can be operated at any angle.
There are a few CCS designs on my web page that will work for 120ma or more. The mosfets used are good for 50 watt dissipation so are much tougher than the DN2540's 15 watt rating. I also sell the boards to build the CCS's on. There are 2 new versions that have not been added to the web page yet. A new battery biased CCS that uses 27A 12V alkaline batteries that snap into battery clips instead of the lithium cells that solder into the board. And a small footprint CCS (same as the battery biased CCS's) that self-biases using the voltage reference from the Rev 5 CCS.
Here is a picture of the newest CCS's in the family
An externally hosted image should be here but it was not working when we last tested it.
Gary
Gary P's DIY page
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.