Aim-TTi (Thurlby Thandar) EL303R PSU teardown 2021

[SerjP]

I bought a new single-channel 30V 3A linear PSU Aim-TTi EL303R.
Links: manufacturer EL-R Series, PDF data sheet

I have a couple of questions on different aspects. I will post them later one at a time.

Links to similar older threads I found so far:
Hello (and the insides of a TTI / Thurlby Thandar PL330 power supply)
Thurlby Thandar... not a happy bunny
Thurlby Thandar QL355 Power Supply
Thurlby Thandar QL355 Power Supply - Pt. 2; the innards
Thurbly Thandar Instruments (TTi) PSU quality ?
TTi PL303QMD Power Supply, Fan noise?

[SerjP]

Does someone have a schematic diagram or service manual for this PSU?
I tried to search it, but can't find a good one. Found that: "Models EL303 and EL155 are covered by a separate manual 48511-0920"

I asked TTi for a diagram via tech query and they send me two single page PDFs. What they sent to me contains partial diagram, which is useful but have some important blocks missing. Like pre-regulator circuit is missed. Yes this linear PSU have zero phase detection IC and pre-regulator before main cap to reduce power losses. I partially reconstructed and draw the schematic to find it out.

So if someone can share the service manual 48511-0920 or schematic it would be great.

[SerjP]

Main transformer have no insulating washer (shoulder washer) at core holes. Is the some new physics nowadays? I thought the core should be electrically isolated from bolts via insulation inserts around bolts.

Also, I have not seen before the welding of the magnetic core as in the second photo.

[Kleinstein]

A weldind strip or more around the core is not so rare. It is Ok, as the electircal connection is only locally and no close loop from this. Even multiple such welded areas are possible.

Ideally the screw should be isolated, but the voltage is relatively low and thus only a little turns voltage. Isolated washers at 1 side would be nice, isolation around the bolts not absolutely needed.

[wraper]

Main transformer have no insulating washer (shoulder washer) at core holes. Is the some new physics nowadays? I thought the core should be electrically isolated from bolts via insulation inserts around bolts.

IME insulated screw is way less common.

Also, I have not seen before the welding of the magnetic core as in the second photo.

I've seen plenty of that. And it's a sign that transformer should be very silent.

[SerjP]

A weldind strip or more around the core is not so rare. It is Ok, as the electircal connection is only locally and no close loop from this. Even multiple such welded areas are possible.

Ideally the screw should be isolated, but the voltage is relatively low and thus only a little turns voltage. Isolated washers at 1 side would be nice, isolation around the bolts not absolutely needed.

IME insulated screw is way less common.

Also, I have not seen before the welding of the magnetic core as in the second photo.

I've seen plenty of that. And it's a sign that transformer should be very silent.

Nice to know, thank you!
Transformer is completely silent at idle. But it generates some noise at any load higher than 20-30mA. I expect it is due to the pre-regulator circuit working in switch mode.

[SerjP]

more photo

I expected to see a better quality main cap. Lets hope it won't become the point of failure in the next 10 years.

[SerjP]

A little rainbow on the forum to cheer up. I don't know what is the manufacturing process at TTi but just looking at this rainbow make me a bit happier.
For those pragmatic: actually I tried to capture the Jamicon cap.

[SerjP]

Control board.
Pre-regulator is based on zero phase detector TCA785P. It keeps the main cap charged to voltage around 7v higher than the PSU output voltage. So that the path-through power transistors dissipates no more than 7V * I watts of heat. The maximum voltage I measured on the cap is 37V. But once the PSU output is set to 1V for example and load get connected, the voltage on the cap dropped to around 8V.

[SerjP]

Some calculations on power dissipation.


And temperature on the key elements during the load test. Measured with Fluke 87V + 80BK kit probe.

[SerjP]

PSU meter accuracy. Everything is well within the spec.
To get a more precise error numbers I follow the next approach. Set some value on PSU meter like 3.00v then slowly turn the fine knob to find a lowest possible knob position while keeping the PSU meter voltage displayed as of 3.00V. Measured output with Fluke 87V and write it down to 'min' column. Then slowly increase output voltage up to the point when PSU meter is going to roll over to the 3.01V. And write it down to 'max' column. So that I can capture min error and max error having 3.00 displayed on PSU screen.

There are couple of control measurements on the right in the table made with Uni-T to just verify the fluke's values.

[SerjP]

A weldind strip or more around the core is not so rare. It is Ok, as the electircal connection is only locally and no close loop from this. Even multiple such welded areas are possible.

Ideally the screw should be isolated, but the voltage is relatively low and thus only a little turns voltage. Isolated washers at 1 side would be nice, isolation around the bolts not absolutely needed.

The core temperature is 39.2C at 200mA load (room temp is 25C). It looks like not everything is perfect in terms of Eddy current. 200mA is less than 10% of PSU capacity.