Products > Test Equipment

Test Equipment Anonymous (TEA) group therapy thread

<< < (24089/25808) > >>

m k:

--- Quote from: bd139 on May 23, 2022, 09:48:01 am ---
--- Quote from: tautech on May 23, 2022, 09:37:02 am ---Yep the Tek prereg SMPS is pretty well behaved except when it isn't.
IIRC didn't you and I discuss MOSFET gate Miller capacitance some while back which is the trap when repairing the Tek prereg as often what looks like a compatible MOSFET for their repair has too much gate capacitance and the switching is troublesome as the driver hasn't sufficient grunt to overcome the gate capacitance.

--- End quote ---

I do believe we did. Another one of the infernal constraints of the universe come to bugger up our engineering attempts.

The killer was if you don't drive it hard enough and quick enough, then it exceeds the SOA while temporarily in the linear region resulting in what sounds like it firing a shotgun at you. I have ear protectors and safety glasses when dealing with SMPS now  :-DD

I think on the 22xx supplies it's not actually sourcing electrons into the gate that was the problem though which came from the TL494, it was getting rid of them afterwards hence the 2n2907 sitting in front of it...



This is of course why it blew the 2n2907, the diode and the MOSFET during the failure cycle  :-DD

--- End quote ---

Or you finally don't have an OC situation since the faulty OVP crowbar is putting the thing to OC situation.

tautech:

--- Quote from: bd139 on May 23, 2022, 09:48:01 am ---
--- Quote from: tautech on May 23, 2022, 09:37:02 am ---Yep the Tek prereg SMPS is pretty well behaved except when it isn't.
IIRC didn't you and I discuss MOSFET gate Miller capacitance some while back which is the trap when repairing the Tek prereg as often what looks like a compatible MOSFET for their repair has too much gate capacitance and the switching is troublesome as the driver hasn't sufficient grunt to overcome the gate capacitance.

--- End quote ---

I do believe we did. Another one of the infernal constraints of the universe come to bugger up our engineering attempts.

The killer was if you don't drive it hard enough and quick enough, then it exceeds the SOA while temporarily in the linear region resulting in what sounds like it firing a shotgun at you. I have ear protectors and safety glasses when dealing with SMPS now  :-DD

I think on the 22xx supplies it's not actually sourcing electrons into the gate that was the problem though which came from the TL494, it was getting rid of them afterwards hence the 2n2907 sitting in front of it...



This is of course why it blew the 2n2907, the diode and the MOSFET during the failure cycle  :-DD

--- End quote ---
No not that one but a Tek repair thread Vince did where member Per Hansson gave us lessons about Miller capacitance:
https://www.eevblog.com/forum/repair/tektronix-2215-scope-repair/msg1824401/#msg1824401
Well worth a read.

Unfortunately also the thread where helpful member oldway crossed swords with Simon and despite it being a private slanging match oldway is with us no more.  :(

bd139:

--- Quote from: tautech on May 23, 2022, 10:26:56 am ---
--- Quote from: bd139 on May 23, 2022, 09:48:01 am ---
--- Quote from: tautech on May 23, 2022, 09:37:02 am ---Yep the Tek prereg SMPS is pretty well behaved except when it isn't.
IIRC didn't you and I discuss MOSFET gate Miller capacitance some while back which is the trap when repairing the Tek prereg as often what looks like a compatible MOSFET for their repair has too much gate capacitance and the switching is troublesome as the driver hasn't sufficient grunt to overcome the gate capacitance.

--- End quote ---

I do believe we did. Another one of the infernal constraints of the universe come to bugger up our engineering attempts.

The killer was if you don't drive it hard enough and quick enough, then it exceeds the SOA while temporarily in the linear region resulting in what sounds like it firing a shotgun at you. I have ear protectors and safety glasses when dealing with SMPS now  :-DD

I think on the 22xx supplies it's not actually sourcing electrons into the gate that was the problem though which came from the TL494, it was getting rid of them afterwards hence the 2n2907 sitting in front of it...



This is of course why it blew the 2n2907, the diode and the MOSFET during the failure cycle  :-DD

--- End quote ---
No not that one but a Tek repair thread Vince did where member Per Hansson gave us lessons about Miller capacitance:
https://www.eevblog.com/forum/repair/tektronix-2215-scope-repair/msg1824401/#msg1824401
Well worth a read.

Unfortunately also the thread where helpful member oldway crossed swords with Simon and despite it being a private slanging match oldway is with us no more.  :(

--- End quote ---

Ah yes - miller plateau is part of the linear region transition i.e. when it's not on or off. Objective is to reduce dT there and eliminate that as quickly as possible by sourcing and sinking more electrons per second. The important thing is really gate charge there: the lower the better!  I notice another MOSFET appeared in there which is nice and all that but you tend to have to grab whatever is cheap that will work. When I did the change I optimised for gate charge (related to integral of current into and out of the gate) and SOA rather than gate capacitance. The problem with the Tek design was the SOA rather than the gate capacitance which I ran in LTspice against a few MOSFETs and there was marginal difference. The IRF820 had a higher pulse current and a little more voltage headroom but sacrificed a little bit of gate capacitance and not much gate charge.

Saskia:
Did have some time to finally frickledoodle that nomoreprivatecustomersexAgilent PSU...
At least the Datatecwechargeyoutosendyouthemissingpartssothatyoucanuseyourpurchase module apparently works.
Fluke 179 measured 35V output sharp.

Cerebus:

--- Quote from: bd139 on May 23, 2022, 08:02:27 am ---Edit: oh meant to mention something grim I saw on the way to Dunstable. So there's an M&S at Acton on the North Circular. Every single sandwich was removed from the shelves due to Salmonella. Big sign up saying what had been taken off the shelves :scared:

--- End quote ---

"There's food poisoning and then there's M&S food poisoning."*

* For those not exposed to UK media: Parody of M&S's current ad campaign slogan.

Navigation

[0] Message Index

[#] Next page

[*] Previous page

There was an error while thanking
Thanking...
Go to full version