Back in 2002 I purchased an Elenco F-1300 universal counter for US$220. The unit offered a high-resolution frequency function below 35 kHz that I thought would be useful for audio work. Instead of measuring the number of input cycles during a 0.1 or 1 second gate time, the F-1300 measures the number of cycles of the master clock during one cycle of the input and then calculates the frequency. Thus for a 1 kHz input you could read 1000.000 Hz with a fast update of the display.
The only issue I have with the stock F-1300 involves the stability of its crystal-controlled timebase. Perhaps the accuracy spec of +/- 10ppm/year with a warm-up time of 30 minutes (I measured close to 60 minutes to final equilibrium) is acceptable for some work. However, I felt I could do better. The timebase circuit incorporates a 10 MHz crystal with gain provided by an inverter from a 74HCU04. I thought the counter's performance could be improved by replacing the timebase circuitry with a temperature-controlled crystal oscillator (TCXO).
There are many possible TCXOs that one could use. I settled on a $20 unit I purchased on eBay from a Chinese company called LogiFind - stores.ebay.com/logifind/. Listed as an "Ultra precision Crystal Oscillator" the spec sheet said that precision was +/- 1 ppm with a 1 ppm/year max ageing rate. The supply current spec of 20 mA max shouldn't overtax the F-1300's regulated +5V supply. The oscillator comes in a standard metal DIP can with four pins. Pin 1 is no-connect so there is no possibility of voltage-controlled frequency trimming. Shipping to the USA was much faster than month or so quoted during the eBay purchase.
To mount the new TCXO I first removed the old timebase components - X32, R29, C34, C35, C36 and C361. The crystal and trimmer are thru-hole and have to be unsoldered from the board bottom which is easy to get to. Prior to mounting the TCXO I soldered a 0.047 uF decoupling capacitor between pins 7 and 14, about halfway up the can leads.
TCXO pin 8 solders to the C35 pad that goes to the R29 pad and then pin 13 of the 74HCU04 (U30). Pin 7 goes to the ground side of surface mount capacitor next to C73. Pin 14 touches the board near a via hole so I first insulated it with a bit of electrical tape on the board. Then I ran a wire from the pin to the +5V pad between C362 and C363. Finally to secure the TCXO I ran a thin cable-tie through the C36 trimmer hole, around the TCXO and back down the hole. See the photo.
I am very happy with the modification. Now after the counter powers up I get an instant reading of 10.000000 from a rubidium standard. The only down-side to this project is that the F-1300 has been discontinued by Elenco. You will occasionally see one for sale on eBay but I have not discovered any other source. But if you are able to obtain one, perhaps this mod will prove useful.
The only issue I have with the stock F-1300 involves the stability of its crystal-controlled timebase. Perhaps the accuracy spec of +/- 10ppm/year with a warm-up time of 30 minutes (I measured close to 60 minutes to final equilibrium) is acceptable for some work. However, I felt I could do better. The timebase circuit incorporates a 10 MHz crystal with gain provided by an inverter from a 74HCU04. I thought the counter's performance could be improved by replacing the timebase circuitry with a temperature-controlled crystal oscillator (TCXO).
There are many possible TCXOs that one could use. I settled on a $20 unit I purchased on eBay from a Chinese company called LogiFind - stores.ebay.com/logifind/. Listed as an "Ultra precision Crystal Oscillator" the spec sheet said that precision was +/- 1 ppm with a 1 ppm/year max ageing rate. The supply current spec of 20 mA max shouldn't overtax the F-1300's regulated +5V supply. The oscillator comes in a standard metal DIP can with four pins. Pin 1 is no-connect so there is no possibility of voltage-controlled frequency trimming. Shipping to the USA was much faster than month or so quoted during the eBay purchase.
To mount the new TCXO I first removed the old timebase components - X32, R29, C34, C35, C36 and C361. The crystal and trimmer are thru-hole and have to be unsoldered from the board bottom which is easy to get to. Prior to mounting the TCXO I soldered a 0.047 uF decoupling capacitor between pins 7 and 14, about halfway up the can leads.
TCXO pin 8 solders to the C35 pad that goes to the R29 pad and then pin 13 of the 74HCU04 (U30). Pin 7 goes to the ground side of surface mount capacitor next to C73. Pin 14 touches the board near a via hole so I first insulated it with a bit of electrical tape on the board. Then I ran a wire from the pin to the +5V pad between C362 and C363. Finally to secure the TCXO I ran a thin cable-tie through the C36 trimmer hole, around the TCXO and back down the hole. See the photo.
I am very happy with the modification. Now after the counter powers up I get an instant reading of 10.000000 from a rubidium standard. The only down-side to this project is that the F-1300 has been discontinued by Elenco. You will occasionally see one for sale on eBay but I have not discovered any other source. But if you are able to obtain one, perhaps this mod will prove useful.
Attachments
Me too
Thanks for the post! I too have an old Elenco F-1300 and wanted better accuracy than the XO provides. I bought a high accuracy (claimed 0.1ppm, actual ~0.2ppm) TCXO on eBay for $10. I mounted the TCXO to the enclosure deadbug style with hot glue and followed your directions and voila...a much more accurate counter! If I were doing this again, I'd add an SPDT switch and external BNC connector to the back panel, and use some thin coax to give the counter a switchable internal TCXO/external reference so I could use it with a lab standard.
The only caveat for others considering this counter is that the F-1300 only has 7+1 digits of resolution so at 10MHz, you have 1Hz resolution (0.1ppm) and if you are looking at VHF/UHF frequencies, your resolution is 100Hz. That said, the gate is pretty quick and with the TCXO, the counter is reasonably accurate too.
If you're doing the mod: 1) Disconnect power!!! 2) unscrew the 4 screws in the black feet at the bottom 3) Lift top cover off (the back panel sits in the enclosure and can remain attached to the PCB) 4) use hot-air to remove the parts specified in the OP 5) Install the TCXO as shown and hot-glue in place. 5) reassembled and test!
Thanks for the post! I too have an old Elenco F-1300 and wanted better accuracy than the XO provides. I bought a high accuracy (claimed 0.1ppm, actual ~0.2ppm) TCXO on eBay for $10. I mounted the TCXO to the enclosure deadbug style with hot glue and followed your directions and voila...a much more accurate counter! If I were doing this again, I'd add an SPDT switch and external BNC connector to the back panel, and use some thin coax to give the counter a switchable internal TCXO/external reference so I could use it with a lab standard.
The only caveat for others considering this counter is that the F-1300 only has 7+1 digits of resolution so at 10MHz, you have 1Hz resolution (0.1ppm) and if you are looking at VHF/UHF frequencies, your resolution is 100Hz. That said, the gate is pretty quick and with the TCXO, the counter is reasonably accurate too.
If you're doing the mod: 1) Disconnect power!!! 2) unscrew the 4 screws in the black feet at the bottom 3) Lift top cover off (the back panel sits in the enclosure and can remain attached to the PCB) 4) use hot-air to remove the parts specified in the OP 5) Install the TCXO as shown and hot-glue in place. 5) reassembled and test!
Attachments
- Status
- Not open for further replies.