Bio-Rad fixes our NGC

Bio-Rad's logo on our NGC FPLC device

The best way to keep your FPLC device in good shape is to have a maintenance contract. Although we have been able to get money from our university to buy a top-of-the-line FPLC twice in the last 30 years, getting money for a service contract is much more difficult. One of the many reasons is that most grant periods are shorter than service contracts, which only make sense if you make them over several years.

Minor trouble started more than a year ago
Lacking such maintenance contracts, our Äkta Explorer in the Medical Faculty and our Bio-Rad NGC in the Faculty of Pharmacy are perhaps not in the best shape they could be. In addition, they are used for teaching purposes and are accessible (after training) to any researcher who needs them. Last summer, we had problems with our NGC's UV module: It did not respond to the status query. Consequently, it was difficult to estimate when we would need to replace the lamps. We ordered service, and I think one lamp had already failed when the service occurred a few weeks later. However, the service engineer could not answer our question as to why the lamp status query failed and - more importantly - why this had not triggered any error messages when the lamp calibration happened during booting. That was a bit unsatisfactory. However, I have been working in an old-fashioned car repair shop when electronic diagnostics was just being introduced. Therefore, I know the change of attitude. Today, car repair often comes down to “The diagnostics did not tell us what is broken. But let's try to change the (INSERT PART NAME HERE) and see whether things are improving".

We need at least one FPLC for teaching
Twice a year, we teach students how to use the machine (DPDR-305 in teaching period 2, and PROV-004 in teaching period 3). On top of the (admittedly minor) UV issue, two more serious problems have been cropping up, which make the use of the device for teaching nearly impossible. We have three problems with increasing severity that we need to solve by December 4th, when our next practical course will start.

  1. This weird problem with the UV readings started in the summer of 2022. Nobody seems to have an explanation. The problem is not reproducible; it seems to happen sporadically. For once, the warm-up time of the lamp lasts sometimes 24+ hours instead of 30 minutes. This has prompted us never to switch the machine off. Of course, the underlying problem remains, and sometimes, the slowly changing baseline appears randomly during a run. That would not be a big problem if the Bio-Rad software would allow for an easy manual adjustment of the baseline. However, I have not found yet such a possibility. As already said, this is a minor problem since it affected perhaps only one in 20 runs. Despite this, every few weeks, someone knocks at my office door, asking for help with an inexplicably misbehaving baseline. My suggestion was always to reboot the machine, which often helped (indicating that this might also be a software issue).
  2. The Bio-Rad NGC becomes nonresponsive to commands from the controlling computer. This often starts as a delayed response (a few seconds) and quickly escalates to response delays of many minutes. The problem always disappears after a reboot. Sometimes, the machine works for weeks without any of these symptoms, but at other times, the problem has cropped up three times in a single day. When contacting our local Bio-Rad representative, the answer was "this might improve when exchanging the motherboard". Our machine has a spinning hard drive, while the newer models use SSDs. The symptoms are indeed reminiscent of a computer that has a failing hard drive, but if it is a software bug, exchanging the motherboard and the hard drive won't help. But it'll set us back a few grands and grants.
  3. In early summer of 2023, in the middle of a run (just before starting the fraction collection), we realized that our fraction collector (Bio-Frac) was unresponsive. Its LCD said something like "Firmware upgrade. Do not switch off". So, we did not switch it off and manually collected our samples from the waste tubing. I waited until the next day for the firmware upgrade to complete but finally decided to switch it off. It never came back online. I am a big fan of pushed software updates. In today's environment, unattended software updates are needed primarily for security and privacy reasons. If a vendor pushes an update, this update should not brick the device. We have a pending grant application that includes the purchase of the NGC fraction collector, but such a collector is a 15k+ investment. We have used the old Bio-Rad standalone fraction collector (which we inherited from the Jorma Keski-Oja laboratory). However, it does not communicate with the NGC. Although it is possible to use it, it is a PITA. Documentation and analysis become a hassle since the chromatogram contains no fraction information. The accuracy of the synchronization of the fractionation start and end depends on the operator's speed (and mostly, we verify peak fractions by measuring OD280nm with our nanodrop).

I have been in contact with the local Bio-Rad support concerning these issues over the last year, but not much has been moving until recently. For almost all newer electronic devices, one can flash new firmware using USB, an SD card or a serial connection. I have done that perhaps 40 times for phones, routers and embedded systems in the early days of the Android OS. However, I don’t have the time to do such things anymore. And my students don’t feel comfortable doing such “risky” operations. My questions to the Bio-Rad technical support were answered by an expert from Central Europe, but apart from the normal advice, which is available from the manual or the internet, nothing ever changed the machine’s deviant behaviour. This summer, we re-activated two Äkta devices from storage: one Äkta Explorer and one Äkta Prime. The Äkta prime worked out of the box (but we still need to find a computer to install the software because it apparently cannot coexist with the Unicorn software on the same computer). The Äkta Explorer is only partially functional since the pumps create only 10% of the flow that the machine indicates. We disassembled the pumps, and the seals appear to be in good condition. This is crucial since I do not know whether they can still be purchased as spare parts from GE Healthcare/Cytiva. GE Healthcare had stopped to guarantee parts availability for the Äkta Explorer about seven years ago. We'll clean the pumps in an ultrasound bath and put them back together, and the device should be in working order soon.

Bio-Rad is stepping up
I don’t know how, but somehow, the information about our trouble had recently finally reached a Bio-Rad representative who cared enough to give me a call. Interestingly, the call came from Sweden. On September 14th, we discussed the situation, and I am happy to report that Bio-Rad agreed to get the machine into working order on a goodwill basis. Taking responsibility for a botched firmware upgrade should be standard for any reputable company (e.g. HP recently had a similar issue with some of their printer models, and they recalled all bricked printers since they could not be fixed remotely). Having been working with expensive research devices, I think I know the reasoning behind this move. Familiarity with a device goes a long way. If you ever need to purchase a device, you will almost always opt for the brand you already know how to use. When we bought the Äkta Avant in 2016, we had been using the Äkta Explorer for the previous 20 years. If you use an iPhone, you probably use it because your previous model was already an iPhone. Looking from that perspective, Bio-Rad’s decision makes a lot of sense. At the moment, the Bio-Rad is our newest device after leaving the Äkta Avant behind in the Faculty of Medicine/Biomedicum when we moved to the Faculty of Pharmacy. If the Bio-Rad NGC worked, we would use it for all our teaching and research work, familiarising many people with its operation. We just received the UV module back from repair, and we are waiting now to get the nonresponsiveness and the botched firmware upgrade fixed, which will likely require the visit of a service technician.

Having two is better than having one
We have, however, already decided that we will invest in a completely new FPLC system in the long run. Even when the NGC had none of the abovementioned problems, people sometimes queued for its use. And the number of users has steadily increased since we joined the faculty. We have used our Äkta Explorer for almost 20 years, and the faculty’s NGC is about ten years old. But buying a completely new top-of-the-line FPLC requires us to allocate 100k or even a bit more. Finding the money took about two years back in 2015. And the funding situation has not improved since. Thus, I fear we again won't be able to pay for a maintenance contract.