G8MNY > TECH 04.05.22 09:32z 92 Lines 4350 Bytes #19 (0) @ WW
BID : 1388_GB7CIP
Subj: Simple Crystal Oven
Sent: 220504/0921Z @:GB7CIP.#32.GBR.EURO #:1388 [Caterham Surrey GBR] $:1388_GB
To : TECH@WW
By G8MNY (Updated Nov 08)
(8 Bit ASCII graphics use code page 437 or 850, Terminal Font)
Here is a design revamped from an old Radcom, I have used it with some success
on a commercial MW Tx to stabilise the frequency. Other uses are for frequency
counters & UHF/SHF Transceivers & Transverters where good frequency accuracy &
stability are important.
Before making an oven try to compensate the Xtal osc with negative temperature
coefficient capacitors instead of existing caps in the osc circuit.
TEMP Net Xtal Measure the frequency e.g. on a SSB Rx. Heat the
50°C ┤ \ ¨ .˙' Xtal with the tip of a soldering iron near by &
40°C ┤ \ ¨ .˙' note the tone change. Substitute a capacitor & heat
30°C ┤ \¨.˙' up that (not melt it) to see if you can get the
20°C ┤ .˙¨\ opposite effect. This can easily half the drift
10°C ┤ .˙' ¨ \Cap without an oven, & if your lucky by far more.
1 2 3 4 5 6 7 8 kHz drift
When you have the best compensation you can get, make an oven around both the
Xtal & the compensation cap.
The aim of the oven is not to cook the Xtal osc & its components, but to
maintain a temperature slightly higher than the hottest running conditions, &
by doing this eliminate frequency drift due to slow warm up & ambient
This is quite simple. But mechanically quite important to get a few things in
the right place.
R1 R2 Vcc The negative temperature coefficient
│ _____│ thermistor TH1 must be mounted above the
│ │/ │ Xtal so that it is the last thing to get
┬ ├──┤NPN │ DARLINGTON hot. The NPN darlington & R2 are placed
Vb │ │\e │/ TAB type below the Xtal, mount the NPN so its tab
1.2V │ └──┤NPN does not touch the Xtal & fill the gaps
| TH1 │\e with heatsink thermal compound.
Here the Xtal is mounted horizontally! ││
TOP..::::::::.. Foam polystyrene ││▒▒▒▒▒▒▒▒▒▒▒▒
::'' <TH1> '':: heat cover PCB ││ <TH1> ▒▒Foam
::: /~~~~~~~~\ ::: outside ││ ┌──────┐ ▒▒cover
::: ( Xtal ) ::: all hot Xtal ││=│ │ ▒▒
::: \________/ ::: components holder││=│______│ ▒▒
::: [NPN](R2) ::: ││ ██████ ▒▒
::..~~~~~ ..:: Thermal past ││ NPN & R2 ▒▒
''::::::::'' between NPN, R2 ││▒▒▒▒▒▒▒▒▒▒▒▒
Xtal & TH1 ││
Over component view Side view
R2 is the main heating component & limits the current & power the heater takes,
it therefore determines the warm up time. Typical value 100ŕ 2W for a 12V rail.
After a short time as everything has started to warm up the NPN darlington then
becomes the main source of heat as the base drive is reduced.
Resistances of R1 & TH1 determine the operate temperature, R1 turning on the
NPN darlington & TH1 turning it off. To find a suitable value for R1, heat up
your thermistor TH1 until it is uncomfortable to hold (50°C) & measure its
Then some simple maths for
(Vcc - Vb) 12V & a darlington gives x9
R1 = ────────── x TH1 that value, should give you
Vb a starting value for R1
Do a test without the cover to check it is all functioning & that the current
goes down (NPN collector voltage goes up) when the lot is hot. Make sure it is
not too hot (increase R1 if it is) that it could melt the foam cover material.
Then make a foam polystyrene heat cover for it all, to dramatically cut down
the power consumption & maintain a steadier temperature. Higher temperature
foam can be made in a mould from expanded foam (cavity sealing/thermal
See also my related tech buls, "Crystal Drift Compensation",
"198kHz Off Air Standard", "Off Air Lock for Ref Osc",
"Comparing Off Air Freq Standards", & "Calibrating Frequency".
Why don't U send an interesting bul?
73 de John G8MNY @ GB7CIP
Read previous mail | Read next mail