Low cost ebay LTZ1000 reference, thoughts?

[NaxFM]

I found this LTZ1000 reference module on ebay. For being a full ready to use module, the price is fairly low, too low I'd say...

What are your thoughts on this? Would you trust this product?

https://www.ebay.com/itm/124538165282?hash=item1cff0d9822:g:dHwAAOSwRFtgCUs-&amdata=enc%3AAQAHAAAA4OVORVY7MbcsaosN6w897MGFhIb1zNhoGDKYj2qQPaL2qX7oYG9TvW4ea8BB9kyopS3qQ9Y%2BWUJdZk6vI4EVgKCGIwhfTX2MSrBqphtm4ZFCR4A05kNoxuCBcw9Kk7ikynk3COTy222OrI5OXH7h9d9Pvl1wKgla%2Fo2z0bZuGy1fuAVc9bCDsisBh2agGXx%2F1UBNKBPelaWn0pNYpwyKVTyGEYTZmpZhyrXK04koPI1ysk4NEtiuPCSLbCSnLvXiTISjghVZVfksm1GyvKz3m4dvzd9EEYkqX8tAMHZb0WOt%7Ctkp%3ABFBM7PW6yPVg

[opa627bm]

it is gone now

[NaxFM]

it is gone now

Sorry, the link i posted was broken.
It works now

[Andreas]

What are your thoughts on this? Would you trust this product?

Hello,

If you have all equipment (LNA, stable measurement devices) you will know it in 1-2 years wether you can trust the device.

There are 2 obvious points from construction.
- the output connector ist not low thermal
- the thermal isolation around the LTZ is obviously missing.

further it seems that there is no output buffer to protect the LTZ against wiring faults.

with best regards

Andreas

[Dr. Frank]

Very bad PCB design, unknown characteristics , like oven temperature (?), was the residual T.C. compensated (?), resistors sufficiently stable (?), no output buffer.

The slots in the PCB, and the lack of low e.m.f. connectors indicate at least, that the creator has not understood the real culprits of the circuit, i.e. how to really build a stable reference (over t, T). Maybe the oven is set to 95°C like in the 3458A, so that would not be useful as a voltage reference.
The price is definitely too high for the BOM, including LTZ1000A, and 5 cheaper BMF resistors, RJ711 is 5ppm/K only. The date code on the LTZ is 9803, if I see it correctly, so that might be salvaged parts at best, or fakes.
The LTZ1000A has very short legs, therefore it has definitely been overheated during de-soldering / soldering, and it's sitting askew, so no really experienced person has assembled the PCB.

Provided it's a "real" LTZ1000A, and no fake, you might encounter a big residual T.C. and a big hysteresis / initial drift directly.
The long term drift can be monitored over a few months, and you'll get an indication how it performs. A good design, assembly and a good selection of oven temperature should give a near zero T.C. and an initial drift of < 1ppm/year, from scratch. It's up to you, what are your expectations or requirements.

In summary, I would not trust this offer at all.

Frank

[dietert1]

At the price, the only alternative is DIY. I doubt whether the average DIY result will be much better than that. One could also buy and then mod the device. I mean adding styrofoam and low thermal binding posts is easy. Adding EMI and overload protection is possible, too. The discussion could be here.
Expecting test reports of buyers..

Regards, Dieter

[NaxFM]

The price is definitely too high for the BOM,

Aside from all the considerations you already made and to which I totally agree, how can you say the price is too high?
It's 170 dollars delivered. Only the LTZ1000A cost like 100 bucks. Add the pcb, resistors, opamp, assembly and shipping and for just 170 dollars I'd say it is barely profitable, assuming of course that the LTZ is new and original...

[Kosmic]

Not sure I would trust a LTZ1000A from an unknown seller in China.

[Jay_Diddy_B]

Hi,

A little zoom in on the picture:



Date on the LTZ1000A:  9803

Date on the LT1013: 9727

recycled parts 

Very cheap socket.

Jay_Diddy_B

[Kleinstein]

It is strange to see the same resistor package for all the resistors, not just the precision ones. This seems to iinclude 2 resistors in the oven part that are not at all cricial. So the resistors seem to be really cheap.

The cut outs look like there are for thermal reasons, but the traces on them are rather wide and thus still quite some heat flow.
The position of the heater control transistor also does not look good and is neither of the 2 good options: a) well separated away from anything critical or b) coupling equally to the whole circuit, e.g with a large heat sink.

So the PCB does not look good as far as I can see. There are better PCB layouts around.

[Dr. Frank]

The price is definitely too high for the BOM,

Aside from all the considerations you already made and to which I totally agree, how can you say the price is too high?
It's 170 dollars delivered. Only the LTZ1000A cost like 100 bucks. Add the pcb, resistors, opamp, assembly and shipping and for just 170 dollars I'd say it is barely profitable, assuming of course that the LTZ is new and original...

The offered part has obviously NOT a new LTZ1000A, but salvaged at best, from 1998.
New ones from digikey are still not available, stock zero, so their listed price of 102$ is obviously not a valid one.
Before the Allocation, they would cost about 50$, and about 40$ for the non-A part, for single parts orders.
Prices of that ballpark you can still find directly at AD, but for 1000EA. I think, 67$ @ 20EA order, but unclear lead time.

All my DIY PCBs with better components and very good stability, as described in the other LTZ - thread, had a BOM-cost of about 100€ each.
I also tried to reduce the cost by using expensive parts only where needed, and a single sided board.
This one has 7 of these RJ711, maybe 9$ each, where only 2 or 3 were needed, but of even better quality (1ppm/K).
So this is an additional argument, that the seller has unnecessarily used too many expensive components, because he obviously has no clue about the crucial parts of the circuit. This we have discussed up and down in this long LTZ thread as well.

Maybe that answers your question, how I can (dare to) say, that this one is too expensive.

Please also take care, that I only spoke about the BOM-cost, from a makers perspective!
Any seller can ask whatever people (who might not be makers themselves) are willing to spend, especially on a finished product.

Frank

[dietert1]

As LTZ1000 or the like isn't available right now on serious dsitribution channels, the offer is a good way to get a working LTZ1000 reference. Of unknown quality. Anyway there seems to be a ppm level calibration printed on top, which is something nice.
Try DIY, e.g. starting with a US$ 136 LTZ1000 from ebay, also China. Then add the board, resistors and some effort until calibration. So there is nothing wrong with the price. Except the seller appears to be "difficult" according to his ebay feedback.

Regards, Dieter

PS: At mouser.com there are 16 evaluation boards with LTZ1000 on stock, each one at € 574. That would be a serious distribution channel.

[sahko123]

I cant tell you about the quality or stability of the circuit but i can confidently say that you should be careful with ltz1000's form china. I have bought one it turned out to be fake. It had a tempco high enough to be a decent thermometer.

[syau]

Mouser accept pre-order of LTZ1000ACH, ordered 2 on 21 Jul, received on 12 Sep. The date code is 2206.

[dietert1]

As of today mouser lists the LTZ1000ACH as:

Auf Bestellung:
    811     verfügbar ab 07-Jul-23
    800     verfügbar ab 23-Jan-24

So the delay already increased to about eight months. For someone with money and an urgent need the best option may be an evaluation kit. Or one can try that ebay offer at 30% of the price.
Or build a JVR instead. Meanwhile i am getting very nice results.

Regards, Dieter

[Kosmic]

Mouser had some LTZ1000A in stock some weeks ago. I bought the last 2 and they are indeed 100$ each now 

[dietert1]

Today mouser shows 698x LTZ1000CH#PBF on stock, unit price is €75.05.

Regards, Dieter