Back in mid-2022 I was given this modem as a gift, and while I wasn’t looking for any V.92 modems I figured I’d document it. While I tend to be a little more V.32 or V.34 inclined, I am open to the idea of V.92 modems doing the work of a V.34 modem. Maybe I just really like the old Rockwell chipsets before they were floated on the stock exchange becoming Conexant Systems.
Externally it is marked as a NetComm Roadster V.92 model AM5698. The power supply required is 7.5V DC centre-positive at a minimum of 450mA which seems to be the standard for the NetComm AM56XX series and the related Banksia Wave SP family of modems.
Internally we can see the PCB is marked with ST309 REV B. Most of the date codes on the board indicate a manufacturing date of mid-late 2004 – which makes this quite a late revision and under 20 years old it came across my desk. 13 capacitors, most if not all were STONE branded. Make note of those capacitor heights. I might have made a slight misstep when replacing them. More on that later.
Oh, what’s that? We’re just jumping straight into the capacitors on this one? Well there you go! Mostly Rubycon YXF series were used, but look what happens when I try to put the case back together…
Dammit! That speaker does not have the clearance it needs. I need the ultra short capacitors. Slap in some shorter ZLG series caps and like magic…
Perfect! Interestingly most of the affected caps were 10uf at 25V but I might have derated them a little to 50V, which in this case made them not quite fit right… 😅 The capacitor at C25 is a 22uF with a 63V rating. Now given it is nearest to the power input it’s probably running at ~7.5V which gives plenty of leeway.
It was of interest it was a bipolar, or non-polar, capacitor. Jon informed me that it was likely because the factory would have just used whatever they had on hand to save on costs or to meet production. When the capacitor was removed, the silkscreen revealed a polarised cap confirming his theory.
Next up we have the ATI (inquiry, information, or interrogation) results. This give us the active firmware version – this wasn’t dumped because I didn’t have much interest in doing it at the time. That said, it might be available through archive.org as an available firmware update. You’ll need to scour the archives to track that down though.
---------- 2022-10-02 20:47:01 +1100: Logging Started ---------- AT&F OK AT&FE0V1S0=0&C1&D2+MR=2;+DR=1;+ER=1;W2 OK ATI0 57600 OK ATI1 255 OK ATI2 OK ATI3 ACF3_V1.702-V90_P21_FSH_LL OK ATI4 NetComm V.92 Modem V1.702SA1 OK ATI5 09 OK ATI6 RCV56DPF-PLL L8571A Rev 47.00/34.00 OK ATI7 OK ATI8 ERROR ATDT1234567890 CONNECT 28800
The capacitors were replaced in this modem in September 2022. The parts listed below are manufacturer part numbers. The formatting isn’t too web friendly, copy the raw text out and paste it into a non-rich text editor (Such as Notepad or TextEdit – in plain text mode) and you’ll probably be fine. 😄
Location Value Voltage Height Width Notes C50 100uF 25V 12mm 6mm 25YXF100MEFC6.3X11 C67 100uF 25V 12mm 6mm As above C25 22uF 63V 12mm 8mm 100YXM22MEFC8X11.5 - Non-polar but had silkscreen for polar, YXM Series, 100V upgrade. C30 4.7uF 50V 6mm 5mm 50YXF4R7MEFC5X11 C65 10uF 25V 6mm 4mm 50YXF10MEFC5X11 - YXF Series, 50V upgrade. C64 10uF 25V 6mm 4mm As above C63 10uF 25V 6mm 4mm As above C62 10uF 25V 6mm 4mm As above C15 10uF 25V 6mm 4mm As above C16 10uF 25V 6mm 4mm As above C17 10uF 25V 6mm 4mm As above C18 10uF 25V 6mm 4mm 25ZLG10MEFC4X7 - Height restricted. C19 10uF 25V 6mm 4mm As above All electrolytic capacitors are radial. Lead spacing is 2-2.5mm on most. Use caution under the speaker as 11/12mm (h) capacitors will not have enough clearance.
As always you can see more pictures in the photo gallery and grab any drivers from the file server. I’ve put an iso image up of the entire driver disc as well but you shouldn’t need much more than the above init string to get dialling.
This modem has been donated to the Australian Computer Museum Society (ACMS).