My grandfather was the founder of Isotemp, he sold it when he retired but I still worked there for until the parent company folded into another company. This is before my time but 96 sounds right as moving into the 2000s no one wanted these larger packages. The inner potentiometer is to adjust the output level which would vary from crystal to crystal. My dad probably designed the dang thing, he's the one who sent me this video so I'm surprised I don't already see him here in the comments.
Since it was soldered inside a metal case already there was no need for a Faraday cage inside. That copper case is a heat spreader to create a uniform heater around the crystal oscillator and crystal. The oven you dismantled isn't all that old as it has surface mnt PCB's.
Looks like the unpopulated parts on the oven control board would be used if you wanted a square wave (TTL) output from the module. Default would be a lower level sine wave. Most OCXO manufacturers offer both options.
I had to repair the HP 10811A double-oven oscillator in a Z3801A GPSDO once and I never thought I'd get into it no less repair it; the first part of the can was soldered. The foam and heating wires made for a very complex take apart. Turned out it was simply a bypass cap all the way down to the bottom. After replacing it I put it back together and sure 'nuff it worked! But what a journey that one was!
Ironically, I am curious about the temperature circuit. The TIP100 is a darlington. So, there's no big resistor, and they have the TIP 100 shorted? I should think that your guess about the comparitor and regulator are spot on. These days, it might be cheaper with a microcontroller, and it would have PID, too.
It's a linear control circuit using an op-amp and whetstone bridge. A small Q and some resistors limit the current, just like an op-amp current limit. MCU's don't have enough resolution of temperature yet to be good controllers. With the 0.062" thick copper can the temp changes about 0.01C/100C.
You had it disassembled enough to change the oscillator. If it is still open, that might make an interesting DIY video. I would be interested in that, because I consider these ovens to be expensive. Changing the oscillator on one with an odd value might be a way to get a thermal unit more cheaply? You would probably need to change some cap values, too, I guess?
I don't know of any company that used 3 MHz, as test equipment was standardized around 10 MHz in the US and UK, etc. Even GPS frequency standards are at 10 MHz.
There was a crystal oven I saw once which had a big glass thermometer sticking out of a hole in the side. I think it may had been an AM exciter crystal.
If you want to compare it to a similar unit, there is a trimble OCXO teardown on the eevblog forum under the topic title " Trimble OCXO 'extreme' teardown "
Going off on a tangent: I know you have a Hakko desoldering tool (and did a video), but I notice you often just grab the solder-sucker. Is this because it isn't worth dragging out the serious tool for just a couple of joints to desolder?
Interesting that the frequency adjust appears to be a variable resistor. Maybe the fine tune of the crystal is done with a reverse biased diode or a varactor ?
I think so, too. At 10:07 one can see a number of sot-23 footprints connected in parallel at the lower side of the oscillator board, of which only one is populated. One of the traces to it comes from a voltage divider and the trace to that one comes from the potentiometer. It is very likely that the single sot-23 component is a voltage controlled capacity and if required, more components could be added in parallel.
As nobody else makes the remark: At 10:26 on the oscillator board at the lower end of the board there is a sot-23 element next to one leg of the quartz, maybe anti-parallel diodes. The conductors are continued to 3 other unpopulated spaces for sot-23s. A nice configuration which is not a hf strip-line at 3MHz. What is that configuration for if it is not an early result of artificial intelligence? :o)
I think I found the answer: The fine adjustment for the frequency is made by a potentiometer that provides a voltage that has to be converted into a capacity. So the sot-23 element could be a capacity diode that has a certain capacity value. To build an universal layout, the mounting space for up to 3 additional diodes in parallel is provided by the configuration.
Hey I designed the thing and I put some comments answering a few questions below. But my comments got lost probably because I put in a link to the specification. Maybe I can add to this comment later.
My grandfather was the founder of Isotemp, he sold it when he retired but I still worked there for until the parent company folded into another company. This is before my time but 96 sounds right as moving into the 2000s no one wanted these larger packages. The inner potentiometer is to adjust the output level which would vary from crystal to crystal. My dad probably designed the dang thing, he's the one who sent me this video so I'm surprised I don't already see him here in the comments.
Since it was soldered inside a metal case already there was no need for a Faraday cage inside. That copper case is a heat spreader to create a uniform heater around the crystal oscillator and crystal. The oven you dismantled isn't all that old as it has surface mnt PCB's.
Wow!
I also wanted to see what's inside an temperature / oven controlled oscillator.
Thanks for the tear-down.
The outer case seems to have a 1996 date. Interestingly complex for a simple output.
Looks like the unpopulated parts on the oven control board would be used if you wanted a square wave (TTL) output from the module. Default would be a lower level sine wave. Most OCXO manufacturers offer both options.
That, or other options: DC (digitally controlled) or VC (Voltage Controlled) atop typical OCXO oscillator electronics.
breaks my heart to see things torn a sunder,...
I had to look it up - the Torx screw drive was patented in 1967.
Isotemp has very detailed reading materials out there about the theory of operations of these things, even about the temperature control.
I had to repair the HP 10811A double-oven oscillator in a Z3801A GPSDO once and I never thought I'd get into it no less repair it; the first part of the can was soldered. The foam and heating wires made for a very complex take apart. Turned out it was simply a bypass cap all the way down to the bottom. After replacing it I put it back together and sure 'nuff it worked! But what a journey that one was!
Been there. Done that. I hooked the supply backwards. Forget what needed replacing, but it worked when back together.
If you trace the oscillator circuit you will find it is similar to the 10811 circuit.
I've always loved them. I grab them out of old equipment that's going away. I'm sure I'll use one of them in a project some day.
I think Torx and SMD components follow a similar history, both from the sixties but gained in popularity in the eighties.
That was built in July of 96. They certainly had SMT and Torx by then.
That is very complex for what it does. Nice tear-down. :)
Fascinating the use of a TIP100 as a heater !...cheers.
I guess it's time to open the crystal can up.
@@andymouse I suppose you could also use a 7805 regulator. They get very hot too. :)
@@frankowalker4662 until the thermal protection kick's in !!!!
@@andymouse I forgot about that. LOL.
The unit was manufactured in the 7th week of 1993. Torx screws were invented in 1967.
It says Date - 9628 right on the top of the unit. Where did you see 9307?
Since there is surface mount components, the torx screws should not be questionable.
Ironically, I am curious about the temperature circuit. The TIP100 is a darlington. So, there's no big resistor, and they have the TIP 100 shorted? I should think that your guess about the comparitor and regulator are spot on. These days, it might be cheaper with a microcontroller, and it would have PID, too.
It's a linear control circuit using an op-amp and whetstone bridge. A small Q and some resistors limit the current, just like an op-amp current limit. MCU's don't have enough resolution of temperature yet to be good controllers. With the 0.062" thick copper can the temp changes about 0.01C/100C.
You had it disassembled enough to change the oscillator. If it is still open, that might make an interesting DIY video. I would be interested in that, because I consider these ovens to be expensive. Changing the oscillator on one with an odd value might be a way to get a thermal unit more cheaply? You would probably need to change some cap values, too, I guess?
Torx - first patent 1967
older than I expected!.
Andy
I don't know of any company that used 3 MHz, as test equipment was standardized around 10 MHz in the US and UK, etc. Even GPS frequency standards are at 10 MHz.
Awesome...cheers !
I first had to buy a set of Torx drivers to work on a ‘72 Porsche, so sometime before then
There was a crystal oven I saw once which had a big glass thermometer sticking out of a hole in the side. I think it may had been an AM exciter crystal.
TORX came out in 1967, by Camcar Textron.
"Fairly new..." I'm guessing the 28th Week of 1996... 🙂 (Date Code on case.)
If you want to compare it to a similar unit, there is a trimble OCXO teardown on the eevblog forum under the topic title " Trimble OCXO 'extreme' teardown "
Going off on a tangent: I know you have a Hakko desoldering tool (and did a video), but I notice you often just grab the solder-sucker. Is this because it isn't worth dragging out the serious tool for just a couple of joints to desolder?
takes too long to heat up for quick stuff
Toryx fastener: developed in 1967.
Good video. Thank God not another laser harp vid.
Interesting that the frequency adjust appears to be a variable resistor. Maybe the fine tune of the crystal is done with a reverse biased diode or a varactor ?
I think so, too. At 10:07 one can see a number of sot-23 footprints connected in parallel at the lower side of the oscillator board, of which only one is populated. One of the traces to it comes from a voltage divider and the trace to that one comes from the potentiometer. It is very likely that the single sot-23 component is a voltage controlled capacity and if required, more components could be added in parallel.
Yes quite a bit of circuitry
The temperature regulator for these works rather well for laser shg crystal heaters. Worked rather well for coherent verdi and avia laser heads. 🤓
As nobody else makes the remark:
At 10:26 on the oscillator board at the lower end of the board there is a sot-23 element next to one leg of the quartz, maybe anti-parallel diodes. The conductors are continued to 3 other unpopulated spaces for sot-23s. A nice configuration which is not a hf strip-line at 3MHz. What is that configuration for if it is not an early result of artificial intelligence? :o)
I think I found the answer: The fine adjustment for the frequency is made by a potentiometer that provides a voltage that has to be converted into a capacity. So the sot-23 element could be a capacity diode that has a certain capacity value. To build an universal layout, the mounting space for up to 3 additional diodes in parallel is provided by the configuration.
@@Manf-ft6zk the SOT-23 are 2 = Transistors, 1 single where are many parallel on the PCB = varactor
T Control needs PI controller. Probably, the D circuitry is not populated on this board.
The crystal itself doesn't look standard, at least not like the two most common cans, I'm curious what's inside :)
a fancy specially selected sc-cut piece of quartz
Hey I designed the thing and I put some comments answering a few questions below. But my comments got lost probably because I put in a link to the specification. Maybe I can add to this comment later.
RUclips loves to delete things. you can find my email on the about page and I can post the links. Good job on the design!
@@IMSAIGuy I found your mail address and sent you mail and some specs. I will comment more on this OCXO and any others you have when I find time.
Bottom end of HF makes me think military
👍👍👏👏....!!
I cringed when you used your fine nippers to clip the case, but it appear to be made of copper, so you are OK!
Second thought, as I watch it looks like steel.
It's probably "mu metal"
Another interesting teardown video showing the significance of perfection of the analogue days.
De VU2RZA